ABDULAZIZ BINAHMED

Oral Carcinoma of the Retromolar Trigone, Maxillary Alveolus and Palate

ACADEMIC DISSERTATIONTo be presented, with the permission of

the Faculty of Medicine of the University of Tampere,for public discussion in the Lecture Room of Finn-Medi 5,Biokatu 12, Tampere, on May 14th, 2009, at 12 o’clock.

UNIVERSITY OF TAMPERE

Considerations in Surgical Management

Reviewed byDocent Kalle Aitasalo University of TurkuFinlandDocent Jyrki TörnwallUniversity of HelsinkiFinland

DistributionBookshop TAJUP.O. Box 61733014 University of TampereFinland

Tel. +358 3 3551 6055Fax +358 3 3551 7685 taju@uta.fiwww.uta.fi/tajuhttp://granum.uta.fi

Cover design byJuha Siro

Acta Universitatis Tamperensis 1406ISBN 978-951-44-7688-4 (print)ISSN-L 1455-1616ISSN 1455-1616

Acta Electronica Universitatis Tamperensis 834ISBN 978-951-44-7689-1 (pdf )ISSN 1456-954Xhttp://acta.uta.fi

Tampereen Yliopistopaino Oy – Juvenes PrintTampere 2009

ACADEMIC DISSERTATIONUniversity of Tampere, REGEA Institute of Regenerative MedicineFinland

Supervised byProfessor George K. B. SándorUniversity of TampereFinlandProfessor Riitta SuuronenUniversity of TampereFinland

To My Wife Munira, & My Daughters Haya & Albandry, For Being the Source of Love, Joy, & Happiness.

Abstract

Cancer of the oral cavity is a most devastating disease that causes a significant disfigurement of the patient with severe morbidity and mortality. It is a challenging disease in its nature and management affecting not only the function but also the cosmesis of the afflicted patients. Despite improvement in the diagnostic and management methods, the prognosis of oral cancer is still unpredictable. This series of works seeks to examine the behaviour and management of oral squamous cell carcinoma at two uncommon sites for which data is lacking; namely the retromolar trigone and the maxillary alveolus / hard palate. Additionally the thesis examines the anatomy of two important neighbouring structures which may be affected by surgical treatment; namely the maxillary artery and the marginal mandibular branch of the facial nerve. Finally the importance of the histological status of the resection margin is examined.

The cohort with squamous cell carcinoma involving the retromolar trigone consisted of 76 cases. The absolute and disease specific survivals at 5 years were 51.4% and 67.7%. In patients treated with surgery, the resection margin status predicted the overall 5- year survival (p= 0.027) with 75% of patients with negative margins surviving 5 years versus a survival of 0% of patients with involved margins. Squamous cell carcinoma of the retromolar trigone has a poor survival for early stage disease. Adequate surgical margins can improve survival.

In analysis of the cohort of 37 patients with tumors involving the maxillary alveolus and hard palate, there was a female predilection. Local recurrence was observed in ten patients with 6 failing at the primary site. The absolute and disease free survival at 5-years was 33% and 62% respectively. The 5-year disease-free survival was 82% for stage I & II and 48% for stage III and IV (p=0.056). No patient treated with radiotherapy as a single treatment modality survived 5 years. Disease free survival for patients treated with surgery, and surgery± radiotherapy, was 69% and 73% at 5 years respectively. (p=0.001). Squamous cell carcinoma of the maxillary alveolus and palate differs from other oral cancers in that the patients are relatively older with a slight female predilection. Treatment with surgery, with or without radiotherapy, appears to improve disease control.

The course of the maxillary artery relative to the lateral pterygoid muscle was dissected in 44 cadaver specimens. In the majority of the cases (30), the artery was found lateral to the lower head of the lateral pterygoid muscle (71% in males and 65% in females). The maxillary artery was found medial to the lower head of the lateral pterygoid muscle in only 14 of the cases (29% in males and 35% in females). No variations in the course of maxillary artery were noted between the two sides and between both sexes. This study shows that the lateral or superficial course of the maxillary artery relative to the lower head of the lateral pterygoid muscle to be more prevalent than the medial or deep course.

The anatomy and function of the facial nerve was studied in 133 patients undergoing neck dissections. When the neck was extended the nerve was displaced in an anterior and downwards direction with the lowest point 1.25 +/- 0.07 cm below the mandible between the posterior and anterior facial veins. When the intent was to preserve the nerve, dysfunction was seen in 16 of 101 dissections (16%).

A total of 425 patients were assessed for adequacy of the margins at the time of resection. After controlling for significant prognostic factors, involved surgical margins increased the risk of death at 5 years by 90%. The status of the surgical margin is an important predictor of outcome. The surgical margin, unlike other prognostic factors is under the direct control of the surgeon.

Keywords: oral carcinoma, maxilla, mandible, retromolar trigone, surgical margins.

Abbreviations

CT Computed Tomography CTX Chemotherapy HPV Human papilloma virus HR Hazard ratio IARC International Agency for Research on Cancer ICD International Classification of Disease IT Immunotherapy MA Maxillary artery MMN Marginal mandibular branch of facial nerve MRI Magnetic Resonance Imaging RMT Retromolar trigone RT Radiation therapy SCC Squamous cell carcinoma SD Standard Deviation TNM Tumor, node and metastasis classification.

Glossary of terms Carcinoma: An epithelial derived group of malignant neoplasms. Facial nerve: The seventh cranial nerve with numerous branches, the terminus of which supply the muscles of facial expression with motor innervation for mimetic function. Five-year disease-free-survival: An analytic term to indicate the success of treatment protocols when comparing different groups of tumors or patients. Hard palate: The horizontal projection of the inferior portion of the maxilla medially. Posteriorly the horizontal process of the palatine bone also joins the maxillary portion to make the hard palate. The entire bony hard plate is covered by attached mucosa and mucoperiosteum on the oral side and respiratory mucosa on the nasal side. Lateral pterygoid muscle: A horizontal bipennate muscle responsible for protrusion of the mandible. Marginal mandibular nerve: One of the terminal branches of the facial nerve supplying the Depressor labii inferioris, the depressor anguli oris, risorius and mentalis muscles. Maxillary alveolus: The most inferior vertical process of the maxilla with sockets to house the maxillary teeth. Maxillary artery: One of two terminal branches of the external carotid artery supplying the midface, masticatory muscles and overlying muscles of facial expression. Resection margin: The limit of the resection, its perimeter, the cut edge of the removed tumor and its surrounding tissue. Retromolar trigone: An ill-defined triangular area in the oral cavity posterior to the upper and lower third molar teeth, with the maxillary tuberosity at its apex. Staging: The categorization of the degree of involvement or extent of a malignant tumor based on clinical examination and or radiographic or magnetic resonance examination. TNM: Acronym for Tumor, Node and Metastasis, for classification system to aid in determining the extent of malignant disease.

List of original papers The thesis is based on the following original articles, which are referred to in the text by the Roman numerals I - V:

I. Binahmed A, Nason RW, Abdoh AA, Sándor GKB (2007). Population-based study of treatment outcomes in squamous cell carcinoma of the retromolar trigone. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology. 104: 662-665. Epub 2007 Aug 29.

II. Binahmed A, Nason RW, Hussain A, Abdoh AA, Sándor GKB (2008). Treatment outcomes in

squamous cell carcinoma of the maxillary alveolus and palate: A population-based study. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology. 105: 750-754. Epub 2008 Feb 23.

III. Hussain A, Binahmed A, Karim A, Sándor GKB (2008). Relationship of the maxillary artery

and lateral pterygoid muscle in a caucasian sample. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology. 105: 31-35. Epub 2007 Aug 29.

IV. Nason RW, Binahmed A, Torchia MG, Thliversis J (2007). Clinical observations of the

anatomy and function of the marginal mandibular nerve. Int J Oral Maxillofac Surg. 36: 712-715. Epub 2007 March 27.

V. Binahmed A, Nason RW, Abdoh AA (2007). The clinical significance of the positive surgical

margin in oral cancer. Oral Oncol. 43: 780-784.

The original publications have been reproduced with the kind permission of the copyright holder.

Contents Abstract Abbreviations ................................................................................................................................................ 5 Glossary of terms ......................................................................................................................................... 6 List of original papers .................................................................................................................................. 7 Contents ........................................................................................................................................................ 8 1 Introduction .............................................................................................................................................. 10 2 Review of the literature .......................................................................................................................... 13 

2.1 Oral Carcinoma ............................................................................................................................... 13 2.1.1 Risk Factors .............................................................................................................................. 14 

2.2 Oral Carcinoma Sites ..................................................................................................................... 16 2.2.1 TNM classification .................................................................................................................... 17 

2.3 Unique aspects of the retromolar trigone, maxillary alveolus and the hard palate ................ 18 2.3.1 Retromolar trigone ................................................................................................................... 18 2.3.2 Maxillary alveolus and hard palate ........................................................................................ 19 

2.4 Anatomic structures at risk ............................................................................................................. 19 2.4.1 Cranial nerve branches and blood vessels .......................................................................... 20 2.4.2 Marginal mandibular branch of the facial nerve .................................................................. 20 2.4.3 Maxillary artery and its branches ........................................................................................... 20 

2.5 Modalities of oral cancer treatment .............................................................................................. 21 2.5.1 Surgical oncology ..................................................................................................................... 21 

2.6 Surgical margins .............................................................................................................................. 23 2.6.1 Margin Classification ............................................................................................................... 23 2.6.2 Frozen sections ........................................................................................................................ 24 2.6.3 Resection planning .................................................................................................................. 25 2.6.4 Post Removal Changes .......................................................................................................... 25 2.6.5 Effect of Margin on Prognosis ................................................................................................ 25 

2.7 Statistical Methods .......................................................................................................................... 26 3 Aims of the study .................................................................................................................................... 27 4 Methods and Materials .......................................................................................................................... 28 

4.1 Subjects ............................................................................................................................................ 28 4.2 Methods ............................................................................................................................................ 28 

4.2.1 Study I Retromolar Trigone Carcinoma ................................................................................ 28 4.2.2 Study II Carcinoma of the Maxillary Alveolus and Hard Palate ......................................... 29 4.2.3 Study III Maxillary Artery ......................................................................................................... 29 4.2.4 Study IV Marginal Mandibular Nerve .................................................................................... 30 4.2.5 Study V Histologic Resection Margins .................................................................................. 30 4.2.6 Statistics .................................................................................................................................... 31 

5 Results ..................................................................................................................................................... 33 5.1 Retromolar Trigone Carcinoma ..................................................................................................... 33 

5.2 Carcinoma of the hard palate and maxillary alveolus ................................................................ 34 5.3 Maxillary Artery ................................................................................................................................ 36 5.4 Marginal Mandibular Nerve ............................................................................................................ 37 5.5 Resection Margin Status ................................................................................................................ 40 

6 Discussion ............................................................................................................................................... 44 6.1 Retromolar Trigone Carcinoma ..................................................................................................... 44 6.2 Carcinoma of the hard palate and maxillary alveolus ................................................................ 45 6.3 Maxillary Artery ................................................................................................................................ 46 6.4 Marginal Mandibular Nerve ............................................................................................................ 48 6.5 Resection Margin Status ................................................................................................................ 49 6.6 Future Perspectives ........................................................................................................................ 51 

7 Summary and conclusions .................................................................................................................... 52 Acknowledgements ................................................................................................................................... 53 8  References .......................................................................................................................................... 55 9 Role in publications ................................................................................................................................ 66 Original Publications .................................................................................................................................. 67 

1 Introduction Cancer of the oral cavity is one of the most devastating diseases to afflict the head and neck causing significant disfigurement of the patient with severe morbidity and mortality. Oral carcinoma is a particularly challenging disease whose nature and management may have devastating effects on the function and cosmesis of patients. Despite all of the improvements in diagnostic and management methods the prognosis of oral cancer remains unpredictable. This series of works seeks to examine the behavior and management of oral squamous cell carcinoma at two uncommon sites for which data is lacking; namely the retromolar trigone and the maxillary alveolus / hard palate. The thesis also examines the anatomy of two important neighboring structures which may be affected by surgical removal of the tumor; namely the maxillary artery and the marginal mandibular branch of the facial nerve. Finally the importance of the histological status of the resection margin is examined. Retromolar Trigone

The retromolar trigone (RMT) is an ill-defined triangular area in the oral cavity posterior to the upper and lower third molar teeth, with the maxillary tuberosity at its apex. It is a relatively uncommon site for oral squamous cell carcinoma. The prognosis for this site tends to be poor because of the advanced stage of disease at the time of presentation (Byers et al 1984, Kowalski et al 1993, Antoniados et al 2003). Squamous cell carcinoma of the RMT has been treated with primary surgery, primary radiation, and with combinations of surgery, radiotherapy, and chemotherapy with varying success (Byers et al 1984, Lo et al cell carcinoma of the RMT is relatively poor, with overall survival ranging from 26% to 80% (Byers et al 1984, Lo et al 1987, Kowalski et al 1993, Antoniados et al 2003, Ayad et al 2005).

There is no agreement on the optimal management of these tumors. Early reports recommended surgical treatment over radiotherapy. Kowalski et al (1993) described their experience with extended “commando” operations in 114 patients with or without postoperative radiation and reported a 5-year absolute survival rate of 55.3%. They concluded that adjunctive radiotherapy improved local and regional control (Kowalski et al 1993). The experience reviewed by Byers et al (1984) showed a recurrence rate of 16% in patients treated with radiation therapy alone and 18% in patients treated with postoperative radiation. The 5-year absolute survival was 26% (Byers et al 1984). In another report from the same institution, Lo, Fletcher & Byers (1987) reviewed experience with irradiation of the anterior faucial pillar and the RMT. The 5-year determinant survival was 83% (Lo et al 1987). In a more recent study Hao et al (2006) reported a 5-year actuarial survival of 60.6% in 50 patients treated with surgery, with 24 receiving adjuvant therapy. This thesis tries to provide further information on SCC of the RMT. Maxillary alveolus and hard palate

Another uncommon tumor, squamous cell carcinoma (SCC) of the hard palate and maxillary alveolar ridge could be considered as a single site since the two areas are adjacent anatomically and share similar clinical presentations and management. Squamous cell carcinoma of the maxillary alveolus and hard palate is relatively uncommon in Western Society (Bhansali 1961, Krolls & Hoffman 1976). Petruzzelli & Mayers (1994) noted that neoplasias, both SCC and non-SCC, of the upper jaw constitute only 2% of all head and neck malignancies and 10% of oral cancers. The lesions appear to be considerably more common

in India, accounting for 40% to 55% of the oropharyngeal cancers (Ramulu & Reddy 1972, Reddy 1974). The relatively low numbers of these tumors is the most likely reason that these lesions are often grouped and reported together with other sites such as the mandibular alveolus and soft palate or combined with salivary gland tumors. It is the feeling of this research group that these tumors should be assessed as a site independent from the mandibular alveolus and soft palate.

The anatomical characteristics of the maxillary alveolar ridge and hard palate are similar as together they form the roof of the oral cavity and the floor of the nasal cavity and maxillary sinus. Full-thickness resection of the structures involved by tumor at this site will result in oronasal and oral-antral communication, a problem unique to this site. This thesis attempts to provide further information on SCC at this site. Additionally management of maxillary tumors requires a thorough understanding of the anatomical structures in this area especially the anatomical variation of the Maxillary Artery (MA) which can be a major cause of bleeding in the surgical management of tumors at this site. The maxillary artery

The MA arises as one of the two terminal branches of the external carotid artery behind the neck of the mandibular condyle, deep to or within the substance of the parotid gland (Allen et al 1974, Coleman & Smith 1982). From its origin, the course of the maxillary artery is divided into 3 parts. The first or “mandibular” part of the MA runs nearly horizontal and deep to the neck of the mandibular condyle. It is superficial to the sphenomandibular ligament during its course towards the lower border of the lower head of lateral pterygoid muscle in the infratemporal fossa. The second or “pterygoid” part of the MA runs obliquely forward either superficial or deep to the lower head of the lateral pterygoid muscle. At this site, it is embedded into the pterygoid venous plexus. Finally, the third or “pterygopalatine” part of the MA disappears from view by diving between the two heads of lateral pterygoid muscle to reach the pterygopalataine fossa via the pterygomaxillary fissure. There, it gives off several branches that accompany the corresponding branches of the maxillary nerve. The variable course of the second part of the maxillary artery has stimulated considerable discussion and a number of reports (Krizan 1960, Fisch & Pillsbury 1997, Krizan et al 1960). Lauber (1901) found only 16 arteries (8.5%) out of 200 cases to be superficial to the lower head of the lateral pterygoid muscle. However, the superficial position of the maxillary artery seemed to occur in about 54%-70% of cases in Caucasians (Lurje 1946, Krizan 1960) compared to 69% in Africans (Laskar et al 1951)

Understanding the anatomy of the human MA and its branches in the infratemporal fossa is a prerequisite for many surgical and clinical procedures. Tumors involving the cavernous sinus and the floor of the middle cranial fossa, may extend into the infratemporal fossa. Other tumors originating in the nasopharynx, the maxillary sinus, the maxilla, maxillary alveolus, hard palate, retromolar trigone and in the parotid gland can also invade the infratemporal fossa and involve the MA. This thesis tries to delineate the anatomy of the MA to facilitate surgery in this area. Marginal mandibular branch of the facial nerve

The management of the cervical lymph nodes in treating oral cancers plays a major role in control of the disease as well as survival of the patients. It is not surprising that the presence of clinically positive lymph nodes at the time of presentation is probably the single most important factor in determining outcome and prognosis. In general patients presenting with neck node metastases do half as well as patients who present with a primary tumor only. The attempt to resect potentially involved cervical lymph nodes may cause injury to the marginal mandibular branch of the facial nerve (MMN) with functional and cosmetic deformity.

Injury to the MMN in the course of a neck dissection has been considered an acceptable outcome by some (Moffat & Ramsden 1977, Hald & Andreassen 1994) and the incidence of nerve dysfunction consequently in this context has not been reported. This is unfortunate in terms of the quality of patient care and the quality of life after resection as damage to this nerve results in a significant deformity.

The deformity is the result of interruption of the nerve fibres to the depressor anguli oris and the depressor labii inferioris that results in flattening and inversion of the ipsilateral lower lip and the inability to move the lip downward and laterally. The defect may not be apparent with the face in repose; it is certainly evident upon animation of the face. Elevation of the affected lower lip is evident when the patient smiles and the asymmetry is most obvious when the patient cries – the ‘asymmetric crying facies’ (Moffat & Ramsden 1977, Tulley et al 2000).

The reported incidence of MMN injury ranges from 0 to 20% following submandibular gland removal (Smith et al 1994). Knowledge of the course and anatomic relations of the MMN in the upper neck is important in order to avoid injury. The nerve is usually described relative to the lower border of the mandible and the facial vessels.

The seminal work on the surgical anatomy of the MMN is based on Dingman and Grabb’s study of 100 embalmed cadaver facial halves (Dingman & Grabb 1962). They described two or more rami of this branch of the facial nerve at the level of the angle of the mandible. In 98% of specimens the nerve passed downward and forward over the surface of the posterior facial vein and more anteriorly the nerve crossed the anterior surface of the anterior facial vein in 100% of specimens (Dingman & Grabb 1962). Immediately anterior to the facial vein the branches of the mandibular ramus passed superficial, deep or on both surfaces of the facial artery. The nerve tended to be superficial to the artery (Dingman & Grabb 1962). It was found that posterior to the facial artery the marginal mandibular branch of the facial nerve ran above the inferior border of the mandible in 81% of specimens. In the other 19% the nerve or one or more of its branches ran in an arc 1 cm or less below the inferior border of the mandible (Dingman & Grabb 1962). In the region anterior to the facial artery, all branches of the MMN were above the inferior border of the mandible (Dingman & Grabb 1962).

These observations implied that incisions placed 2 cm below the lower border of the mandible would avoid injury to the nerve in all instances. Clinical observations, however, suggest that the nerve is consistently below the inferior border of the mandible. Baker & Conley noted that in individuals with lax and atrophic tissue, the nerve could be as low as 3 to4 cm below this point (Baker & Conley 1979). This thesis studies the clinical location of this nerve at the time of surgery to try to clarify the anatomic relationships of the MMN relative to other anatomic structures without the potential artefacts of embalming. This would help surgeons avoid injury to this important nerve, potentially improving the functional outcomes of ablative tumor surgery in this area. Histological status of the resection margins

Despite the improvements in surgical and radiotherapeutic technique, intraoral squamous cell carcinoma has a relatively unfavorable prognosis with an overall five-year survival rate of 35 – 50%. The survival rate has regrettably remained virtually unchanged over the past three decades. The increase in the incidence of oral cancer further accentuates the mortality from the cancer.

Several parameters have been adapted and applied by clinicians to evaluate the prognosis of oral cancer. These parameters can be divided into epidemiological parameters which include the age, sex, race, alcohol and tobacco intake and comorbidities; clinical parameters which include the TNM classification and stage and the site of the primary tumor; and histological parameters which include the marginal status, the perineural or perivascular invasion, histopathological grading and tumor thickness and extracapsular spread.

There is no consensus as to how much normal tissue should be removed around the tumor in order to reduce the risk of local recurrence and improved survival. The ancient anatomist Galen suggested that “when excising a malignant tumor one should make accurate incisions surrounding the tumor so as not to leave a single root”. This statement made two millennia ago remains one of the basic principles of oncological surgery. If a cure is to be expected for localized cancer, complete excision must be performed with adequate resection margins.

There is little doubt that a gross residual tumor will effect local control and nearly always increase mortality. Analysis of the literature shows that there is a lack of standardization and precision in defining what is a positive margin, a close margin and a clear margin. There is no prospective controlled study evaluating the effect of positive margins on local recurrence and overall disease free survival (Batsakis 1999).

Data indicates that about 15 to 30% of all oral carcinoma are inadequately excised (Loree & Strong 1990, Batsakis 1999). Some have claimed that the microscopic accuracy of inadequate surgical margin is at least 50% as judged by local recurrence (Batsakis 1999). This false negative assessment is partially due to pathological and surgical difficulties with evaluation of preoperatively irradiated tissue and submucosal extension or skip lesion areas. Another important factor is the miscommunication and lack of close cooperation between surgeons and pathologists as reported (Batsakis 1990).

This thesis also examines the hypothesis that surgical management of squamous cell carcinoma of specific sites in the oral cavity with clear margin improves the prognosis and survival of patients.

2 Review of the literature

2.1 Oral Carcinoma Oral Carcinoma is a potentially disfiguring and devastating disease that robs the patients of their physical appearance and devastates their self esteem. Oral cavity cancers account for about 3% of all cancers diagnosed each year in North America. The ICD codes for oral cancer include ICD9 140, 141, 143-146, 148 and 149, and ICD10 C00-06, C09, C10, C12-14. A 3% incidence is estimated to represent 27,000 newly diagnosed cases in the United States and 3,200 cases in Canada (Figure 1). In the United Kingdom in 2005 there were a total of 4,926 new cases diagnosed with oral cancer (U.K. Cancer Statistics 2009). Slightly more than 10,000 Americans and 1,000 Canadians will die of oral cancer each year (Canadian Cancer Statistics 2004, Jemal et al 2005). Studies from around the globe show that for both sexes combined cancer of the mouth and pharynx ranks sixth overall behind lung, stomach, breast, colon and rectum and cervix uteri in that order (Pisani et al 1993). The incidence ranges from a low of 1.8/100,000 per year to a high of 47/100,000 per year (Johnson et al 1977). In Europe the incidence in males ranges from 9.5 to 23.8 per 100,000 and 3.0 to 4.6 per 100,000 in females (U.K. Cancer Statistics 2009). The highest rates of oral cancer in the world are found in France (Peng et al 2000), the Indian subcontinent, Brazil (Wunsch-Filho & de Camargo 2001) and Central or Eastern Europe. The incidence of tongue cancer in Northern Finland for example is 1.0 to 1.4/100,000 (Kari et al 1997). There are also marked differences between countries in the same geographic regions (Moore et al 2000a, Moore et al 2000b). The incidence of oral cancer increases with age in all parts of the world. In western countries, 98% of the patients are over 40 years of age. In the high prevalence areas of the world many of the patients are less than 35-years-old owing to heavy usage of various forms of tobacco. Furthermore it is now clear that in many Western countries there has been an alarming rise in the incidence of oral cancer during the past two or three decades, particularly among younger men. This trend appears to be continuing (Llewellyn et al 2001). In industrialized countries men are affected two to three times as often as women. The most important risk factors are alcohol and tobacco consumption for intraoral cancer and sun exposure for lip cancer in those who work outdoors. The incidence of tongue and other intraoral cancer for woman can be greater or equal to that of men in high incidence areas such as India where chewing tobacco is also common among women (Shah 2003). There has been a gradual increase of the number of female patients reflected by the change in male to female ratio in Western societies. In 1996 Oliver et al (1996) reported a review of 92 cases of oral squamous cell carcinoma from 1985 – 1992 and found the male to female ratio of 1.5:1.

Figure 1. Appearance of an oral squamous cell carcinoma lesion involving the mandibular anterior alveolus.

2.1.1 Risk Factors

The recognized risk factors of oral carcinoma include tobacco, alcohol, nutritional deficiencies and exposure to viruses. Tobacco

There is absolutely no doubt that on a global scale the use and abuse of tobacco products is a major risk factor of oral cancer. Typically 90% of men and 60% of women with oral carcinomas use tobacco (Vincent & Marchetta 1963, Johnston & Ballantyne 1977, Baker 1993). The incidence rate of oral carcinoma in smokers is six to ten times greater when compared to non-smokers. (Silverman & Griffith 1972). There is strong evidence that active smoking is related to oral cancers. The most comprehensive source of evidence remains the International Agency for Research on Cancer or IARC (IARC 1986, Sasco et al 1986). This evidence is also summarized by the US Surgeon General’s report of 1989. The attributable risk for oral cancer was the highest in males at 92% (Figure 2). Smoking intensity effects have been found to be quantitatively homogeneous across multiple case control studies of lung, bladder, oral

cavity, pancreas and esophagus cancers (Lubin et al 2008). Some literature suggests that pipe and cigars are less risky for oral cancer than cigarettes (Wynder et al 1977) but a very extensive study from northern Italy shows higher risks associated with these practices for cancer of the mouth and esophagus than with cigarettes (Franceschi et al 1992). Reverse oral cigarette smoking has also been shown to carry a high risk for oral cancer development (Reddy 1974) In Scandinavia and North America snuff or smokeless tobacco has been the source of controversy for several years now. Brown et al (Brown et al 1965) described snuff dipper’s cancer in the south eastern United States where the habit of placing snuff in the lower labial sulcus was common. This was the basis of the classical description of verrucous carcinoma by Ackermann (1948) and confirmed by McCoy & Waldron (1981). A study based on combined data of the 1986 US National Mortality Follow Back Survey and the coincident National Health Interview Survey concluded that the use of smokeless tobacco either as a snuff or chewing tobacco does not increase the risk of oral cancer or of cancer of the digestive organs (Sterling et al 1992). In Scandinavia it is now becoming clear that local snuff application is not a major risk factor for oral cancer. Two recent case control studies of the oral cancer cases in Sweden have failed to show an association (Lewin et al 1998, Schildt et al 1998).

Figure 2. Reverse oral cigarette smoking is one habit thought to increase the risk of developing an oral squamous cell carcinoma. Betel quid is a completely different form of smokeless tobacco. It is prepared from areca nut, then cured or sun dried and chopped. These pieces are placed on a leaf of the Piper betel vine and slaked lime is an essential ingredient that is added to the combination to lower the pH and accelerate the release of alkaloid from both tobacco and nut. It is a very common practice in India, Sri Lanka and Pakistan, Bangladesh, Thailand, Cambodia, Malaysian, Singapore, Indonesia, Philippine, Taiwan and China. The carcinogenicity of the betel quad mixture has been clearly established in meta-analysis of 17 published studies by Thomas and Wilson (Thomas & Wilson 1993). Gupta et al (1980) followed up 30,000 individuals over a ten-year period in three areas of India. The annual adjusted incidence of oral cancer was 23 per 100,000 among betel quad tobacco chewers compared with zero in smokers and non-chewers.

Alcohol

It is very difficult to separate the effects of alcohol and tobacco as most heavy alcohol consumers also use tobacco. Nevertheless some cohort and case control studies have found an increased risk of upper aerodigestive tract cancer associated with alcohol drinking in non-smokers (Kato et al 1994). The epidemiological evidence shows that all of alcoholic drinks are dangerous if heavily consumed (Schlecht et al 2001). A Finnish study showed that occupations associated with the highest incidence of oropharyngeal carcinomas were the ones with high consumption of alcohol. Exposure to solvents and possibly pesticides, engine exhaust, textile dust and leather dust may increase the risk of cancer of the mout and pharynx

(Tarvainen et al 2008). Boffetta et al (1992) have shown that the carcinogenic effect of tobacco and alcohol is through direct contact and tends to be site specific in the oral cavity. Tobacco smoking was more closely associated with carcinoma of the soft palate and alcohol was more closely associated with carcinoma of the floor of mouth and tongue. There are several ways in which alcohol is thought to contribute to head and neck cancer by both local and systemic mechanism. The most prominent one is that ethanol increases the permeability of oral mucosa to water and many water-soluble molecules including important carcinogens. The other way that alcohol can contribute to head and neck cancer is through the immediate metabolite of ethanol, acetaldehyde, which has the capability of damaging the cells (Schlecht et al 2001). Alcoholic liver disease is common in heavy drinkers and this reduces the detoxification of active carcinogens (La Vecchia & Tavani 1998). Alcohol is also high in calories, which can suppress appetite in heavy alcohol consumers (Rennie & MacDonald 1984) and cause malnutrition, possibly contributing to oral cancer.

Nutritional Deficiencies

Nutritional deficiencies have been associated with oral cavity carcinoma. Original papers published in 1919 showed a relationship between iron deficiency and increased incidence of upper GI tract cancer in middle-aged women suffering from chronic anaemia. Several animal studies have shown that there is an epithelial cell kinetic change with iron deficiency. Some of the studies have shown an increase of turnover (Rennie & MacDonald 1984) that theoretically increases the risk of mutational error where as other studies have shown a decrease in mutational errors (Ranasinghe et al 1987). Both studies, however, have shown epithelial atrophy and increased cancer risk. Eleven of thirteen case control studies that have examined association between fruit and vegetable consumption and oropharyngeal cancer report a meaningful inverse association (La Vecchia & Tavani 1998, Block et al 1992, Potter & Steinmetz Winn 1995, McLaughlin et al 1988).

Viruses and Other Factors

The understanding of the role of viruses as human carcinogens has progressed in the past two to three decades. Viruses contribute to the multi-step process of carcinogenesis in many human neoplasms. Human papilloma virus (HPV), especially HPV 16, is the most common type associated with not only cervical but also oral and laryngeal cancers (Yeudall 1992, Woods et al 1993, Mao 1995, Syrjänen and Surjänen 1981). In vitro studies show that high-risk HPV types can immortalize primary human oral epithelial cells. However, exposure to tobacco related chemicals were required for these cells to progress to a fully malignant phenotype (Kim et al 1993, Shin et al 1996).

Increased cancer incidence has been documented following organ transplantation. A total of 2,884 renal transplant patients in Finland with a mean follow-up of ten years were found to have an incidence o 0.8% of non cutaneous head and neck cancer (Makitie et al 2008).

2.2 Oral Carcinoma Sites Oral cancer can occur anywhere in the oral cavity. The sites are classically divided into lip, tongue, floor of mouth, cheek, maxillary alveolus, hard palate, soft palate, mandible and retromolar trigone. Cancers of the tonsils and pharynx are not included in oral sites for the purposes of this thesis. The focus of this thesis regarding sites includes squamous cell carcinoma of the retromolar trigone and the maxillary alveolus and hard palate.

2.2.1 TNM classification

Staging of a tumor represents the categorization of the degree of involvement or extent of a malignant tumor based on clinical examination and or radiographic or magnetic resonance examination.The TNM (tumor, node and metastasis) classification was first described by Dr. Pierre Denoix at the Institute Gustaveroussy in France (International Union Against Cancer, 2002). Subsequently the International Union Against Cancer established a special committee on clinical stage classification under the leadership of Dr. Denoix and continued to develop the TNM classification. The American Joint Committee for Staging was organized in 1959 for the purpose of developing a system of clinical staging of cancer. Both organizations combined their efforts in 1988 and agreed upon the TNM classification system which was published at that time (International Union Against Cancer, 2002). There have been multiple modifications of this system since then. The most recent version was published in 2003. The TNM system reflects the anatomical extent of the tumor. The “T” represents the size of the tumor at the primary site (T-stage) emphasizes the size or surface dimension of the tumor. A recognized deficiency of this system is the omission of tumor thickness or depth of infiltration. The “N” represents the involvement of regional cervical lymph nodes and is uniform for the classification of all malignant epithelial tumors of the upper aerodigestive tract (Figure 3). The nodal staging system takes into account the size, multiplicity, and location of metastatic nodes. The presence or absence of distant metastasis as determined by clinical exam and imaging studies, is documented as “M0” or “M1” by routine work up which may include chest radiographs, nuclear medicine scans, CT scans and MRI scans. While oral SCC tends to spread by direct extension, lymphatic spread and has a low incidence of hematologic spread, intramedullary and intradural spinal cord metastasis have been reported (Törnwall et al 2008).

Staging is important for prognostication. The stage of tumor at the time of diagnosis has been related to survival. The overall survival rates are higher in patients with stages I or II cancer are higher than those for stage III (Sargeran et al 2008).

Figure 3. TNM classification relating T or tumor size and M or nodal status to large Roman numerals representing stage I to IV.

2.3 Unique aspects of the retromolar trigone, maxillary alveolus and the hard palate

The retromolar trigone and maxillary alveolus and hard palate are anatomically unique sites of oral cancer. They are poorly documented in the literature.

2.3.1 Retromolar trigone

The Retromolar trigone (RMT) is an ill-defined area posterior to the upper and lower third molar teeth. It is a triangle in shape with the maxillary tuberosity forming the apex, while the base is posterior to the mandibular third molar. The RMT is continuous laterally with the buccal mucosa, medially with the soft palate. Superiorly and inferiorly it blends with the mucoperiosteum of the alveoli (McGregor & McGregor 1986). The RMT is relatively uncommon site for oral squamous cell carcinoma (Figure 4). It is often diagnosed at advanced stage due to their oligosymptomatic early manifestations. In advanced stages pain and trismus are common due to the invasion of the mandibular nerve branches (buccal, lingual, and inferior alveolar) and muscles of mastication. Tumors of the RMT are characterized by early invasion of the mandible and the adjacent buccal mucosa and soft palate.

Figure 4. Oral squamous cell carcinoma involving the retromolar trigone.

2.3.2 Maxillary alveolus and hard palate

The anatomical characteristics of the maxillary alveolar ridge and hard palate are similar as together they form the roof of the oral cavity and the floor of the nasal cavity and maxillary sinuses. Therefore, the clinical presentation, the surgical management and reconstructive considerations, are similar for these sites. From a clinical perspective it is often difficult in advanced tumors to determine the exact epicentre (Figure 5).

Squamous cell carcinoma (SCC) of the hard palate and maxillary alveolar ridge could be considered as one site since the two areas are adjacent anatomically and share similar clinical presentations and management. These are relatively uncommon oral cancers, with a reported incidence ranging from 0.5 % to 2 % (Evan & Shah 1981, New & Hallberg 1994). Many other investigators have even considered SCC of the maxillary and mandibular alveolar ridges as similar entities (Cady & Catlin 1969, Evan & Shah 1981, Soo et al 1988, Gomez et al 2000). However, the behaviour and management of SCC involving the maxillary and mandibular alveolar ridges is different, and therefore the interpretation of their results if combined may be difficult. Other publications have not only compared and contrasted SCC of these subsites but have included other tumors of non-squamous cell origin such as salivary gland neoplasm (Chung et al 1980, Shibuya et al 1984).

Figure 5. Oral squamous cell carcinoma lesion involving the maxillary alveolus and hard palate.

2.4 Anatomic structures at risk While all structures of the head and neck are at risk during tumor removal, damage to blood vessel and even more so damage to cranial nerves may result in significant post operative morbidity for the patient (Dingman & Grabb 1962, Baker & Conley 1979).

2.4.1 Cranial nerve branches and blood vessels

The structures at risk during oncologic therapy of the head and neck include branches of the following nerves: cranial nerves V, VII, IX, X, XI and XII.

Iatrogenic injury is the most frequent cause of sensory disturbances in the distributions of the inferior alveolar and mental nerves (Haskell 1986, Meyer 1990, Carmichael & McGowan 1992, Ventä & Lindqvist 1998) and motor disturbances of the marginal mandibular branch of the facial nerve (Dingman & Grabb 1962).

All branches of the external carotid artery are at risk. When speaking about the resection of an oral cancer the branches at risk specifically include the facial, lingual, and maxillary arteries.

2.4.2 Marginal mandibular branch of the facial nerve

The knowledge of the course and anatomic relations of the marginal mandibular nerve of the facial nerve (MMN) in the upper neck are important in avoiding injury. The nerve is usually described as it relates to the lower border of the mandible and the facial vessels. The seminal work on the surgical anatomy of the marginal mandibular nerve is based on Dingman and Grabb's study (Dingman & Grabb 1962) of 100 embalmed cadaver facial halves. They described 2 or more rami of this branch of the facial nerve at the level of the angle of the mandible. In 98% of specimens the nerve passed downward and forward over the surface of the posterior facial vein and more anteriorly the nerve crossed the anterior surface of the anterior facial vein in 100% of specimens. Immediately anterior to the facial vein the branches of the mandibular ramus passed superficial, deep, or on both surfaces of the facial artery. They tended to be superficial to the artery. They found that posterior to the facial artery the mandibular ramus ran above the inferior border of the mandible in 81% of their specimens and in the other 19% the nerve or one or more of its branches ran in an arc one cm or less below the inferior border of the mandible. Anterior to the facial artery all branches were above the inferior border of the mandible. Dingman & Grabb (1962) made observations and other cadaver dissections (Ziarah et al 1981) imply that incisions placed 2 centimeters below the lower border of the mandible will avoid injury to the nerve in all instances. Clinical observations (Baker & Conley 1979, Nelson et al 1979) suggest that the nerve is consistently below the inferior border of the mandible. Baker and Conley (1979) noted that in individuals with lax and atrophic tissue, the nerve could be as low as 3 to 4 cm below this point.

The MMN is particularly at risk when operating in the upper neck. Injury to the MNN results in a significant cosmetic deformity. The deformity is caused by interruption of nerve fibres to the depressor anguli oris and the depressor labii inferioris that results in flattening and inversion of the ipsilateral lower lip and the inability to move the lip downward and laterally. The defect may not be apparent with the face in repose. Elevation of the affected lower lip is evident when the patient smiles and the asymmetry is most obvious when the patient cries “the asymmetric crying facies" (Moffat & Ramsden 1977, Tulley et al 2000). The reported incidence of marginal mandibular nerve injury ranges from 0 to 20% following submandibular gland removal (Smith et al 1993, Hald & Andreassen 1994, Ichimura et al 1997). Injury to the marginal mandibular nerve in the course of a neck dissection has been considered an acceptable outcome (Hald & Andreassen 1994, Moffat & Ramsden 1977) and the incidence of nerve dysfunction in this context is not reported.

2.4.3 Maxillary artery and its branches

The maxillary artery arises as one of the two terminal branches of the external carotid artery behind the neck of the mandibular condyle, deep to or within the substance of the parotid gland (Fisch 1978, Coleman & Smith 1982). From its origin, the course of the maxillary artery is divided into three parts: The first or “mandibular” part runs nearly horizontal and deep to the neck of the mandibular condyle, and superficial to

the sphenomandibular ligament during its course to the lower border of the lower head of lateral pterygoid muscle in the infratemporal fossa. The second or “pterygoid” part runs obliquely forward either superficial or deep to the lower head of the lateral pterygoid muscle. At this site, it is embedded into the pterygoid venous plexus. Finally, the third or “pterygopalatine” part disappears from view by diving between the two heads of lateral pterygoid muscle to reach the pterygopalataine fossa via the pterygomaxillary fissure. There, it gives off several branches that accompany the corresponding branches of maxillary nerve.

The highly variable course of the second part of maxillary artery has stimulated a considerable number of reports (Krizan 1960, Fisch & Pillsbury 1997). Lauber found only 16 arteries (8.5%) out of 200 cases superficial to the lower head of lateral pterygoid muscle. However, the superficial position of the maxillary artery seems to occur in about 54 to 70% of the cases in Caucasians (Thompson 1891, Lurje 1946, Laskar et al 1951, Krizan 1960, Fisch 1978, Fisch 1982) and in 69% in Africans (Laskar et al 1951). Adachi (1928) stated that the superficial position of the second part of MA occurs with highest frequency in the Japanese population, and referred to it as a “racial difference”. In support of this idea, Laskar and co-workers (Laskar et al 1951) found the artery deep to the muscle in 46% of 147 cases in Caucasians, and in 31% of 61 cases in Africans, and hence, coined the term “racial blood vessel”.

2.5 Modalities of oral cancer treatment Oral cancer therapy may consist of surgical resection, radiation therapy, chemotherapy and immunotherapy.

2.5.1 Surgical oncology

Surgical oncology is the treatment of an oral cancer primarily with surgical resection. Occasionally surgery may also be required to manage a recurrence following primary radiation therapy or as a salvage surgery (Meier et al 1984).

2.5.1.1 Surgical resection

Surgical resection of a tumor involves excising all clinical tumor with a surrounding margin of normal tissue it should bear in mind that tumors presented in three dimensions when treating it. The balance of curing the disease and maintaining the functional and cosmetic aspects is like walking a fine line that requires full knowledge of the clinical anatomy of the area (Baker 1993). Surgery still plays the major treatment modality in most of oral cancer. Surgical resection provides a cost effective and expeditious treatment for this kind of cancer (McCarty & Million 1994). Surgery also has fewer side effects than other modalities of treatments (Jacobs et al 1993), although the late stages of oral cancer still require other modalities of treatment in conjunction with the surgical resection.

There are several approaches to the oral cavity that are used in treating oral cancer. The simplest approach is the per-oral approach that is used in early stage and accessible tumors (Figures 6 and 7). Other more invasive approaches include the check flap with or with out mandibulotomy to have more access to advanced tumors or tumors seated in the posterior aspect of the oral cavity (Baker 1993). More advanced tumors of the maxilla may require a Weber-Ferguson approach to have a full access for resection (McCarty & Million 1994).

Figure 6. Partial mandibulectomy and neck dissection to remove lesion of retromolar trigone.

Figure 7. Neck dissection specimen and mandibulectomy specimens along with margins to check for completeness of tumor removal.

2.5.1.2 Radiation therapy

Radiation therapy (RT) can be used either as a primary modality but more commonly today as an adjunctive treatment to achieve tumor control. Postoperative radiation therapy has an established role in the management of advanced squamous cell carcinoma of the head and neck. Amdur et al (1989) in 1989 and Jacobs et al (1993) reported that the local failure in patients that were treated with surgery with positive margins followed by adjuvant radiation therapy was higher than the patients with negative margins. One important factor influencing local control in patients with positive margins is the dose of postoperative radiotherapy. Zelefsky et al (1993) reported a seven-year local control of 92% in patients who received greater than 60 Gray compared with 44% in patients who received less than 60 Gray showing improved survival. Beitler et al (1998) reported a study of 29 patients with microscopically close or positive margins after curative surgery treated with postoperative external beam radiotherapy to a median dose of 60 Gray followed by iodine-125 permanent implant designed to deliver an accumulative life-time dose of 120 to 160 Gray to the high risk target volume. This treatment strategy resulted in 92% two-year local control.

2.5.1.3 Chemotherapy and immunotherapy

Chemotherapy (CTX) and immunotherapy (IT) are generally adjunctive treatments useful in the management of oral cancer. Their role in management of oral cancer is still in the investigation phase, though CT and IT are used in other head and neck cancers.

2.6 Surgical margins There is no consensus as to how much normal tissue should be removed around a particular tumor in order to reduce the risk of local recurrence and improve the survival. Galen suggested that “when excising a malignant tumor one should make accurate incisions surrounding the tumor so as not to leave a single root” (McCarty & Million 1994). This statement, made two millennia ago, is still one of the basic principles of oncological surgery. If a cure is to be expected for localized cancer, complete excision must be performed with adequate resection margins.

2.6.1 Margin Classification

The definition of positive close and negative surgical margin lacks standardization and certainly precision (Figure 8).

In a recent study by Meier et al (2005) administered a questionnaire to members of the American Head and Neck Society. A total of 1500 surveys were mailed and a total of 476 completed surveys were received. In the definition of clear margin 46% defined a clear margin as more than 5 mm on microscopic examination from the edge of the tumor on permanent section. The second most common response given was that the margin varied according to the clinical situation or that multiple criteria were used. Twelve percent reported that a clear margin was a margin that did not have an ink stain on the tumor edge. Participants were also asked what they considered to be a close margin: 69% defined close as a margin within 5 mm of the tumor on microscopic examination; 21% defined a closed margin as less than one high

powered field; and 9% defined the closed margin with other criteria. When the surgeons were asked if they considered carcinoma in situ or dysplasia at the margin as a positive margin, 83% considered carcinoma in situ as a positive margin while 76% did not consider dysplasia as a positive margin. This study illustrated the lack of standardization in definition of the margin status.

Figure 8. Photomicrograph showing black ink visible on the surgical margin with invasive squamous cell carcinoma immediately adjacent to the ink. This illustrates a positive surgical margin. Hematoxylin and eosin section 200x magnification.

2.6.2 Frozen sections

Intraoperative frozen section analysis of margins is widely employed to assist in complete tumor resection. The technique of frozen section preparation dates to 1818 and has been credited to the work of Deremier (Krubhaar 1937). It has improved over the years. The accuracy of frozen section diagnosis compared with permanent control in head and neck surgery has consistently been reported between 95% and 99%.

Several publications have assessed the accuracy of frozen section margins relative to the adequacy of tumor resection through measurements of local control and disease free survival. Byers et al (1978) reviewed a selected group of 216 patients undergoing surgical treatment of squamous cell carcinoma of the head and neck. Twenty-three percent of this sample size then underwent an intraoperative margin revision due to a positive frozen section study. The local recurrence rate for free margin was 14%, revised margin was 20%, and for the positive margin 80%. In his study Byers et al support the use of frozen section margin evaluation in all cases of squamous cell carcinoma of the head and neck except T4 lesions. Byers et al reported that there is a poor local control rate in these tumors despite the margin status. Ord et al (1997) in a study of 49 patients with oral SCC noted an accuracy rate for frozen section margins of 99% compared with permanent section of the same tissue. However, 70% of patients with a positive final margin on the resection specimen were not detected by frozen section evaluation. A positive final margin included resection margins containing dysplasia, carcinoma in situ, infiltrating carcinoma and margins within 5 mm of the carcinoma. There are two forms of error that can occur in frozen section margin evaluation. The first one is interpretive and the second is the sampling. Interpretive error denotes a failure to correctly identify the tissue present on frozen section slides. The sampling error is mainly related to the surgeon sampling a

non-representative tissue and this has been reported to be mainly at the deep margin (Byers et al 1978). New techniques need to be employed to advance the sampling and intra-operative assessment of the deep margin of squamous cell carcinoma of the oral cavity.

2.6.3 Resection planning

The initial margins are determined by the planned extent of the resection which is dependent on the surgical staging and the balance between achieving disease free status and the sacrifice of life sustaining structures and the ultimate morbidity of their long-term loss (McCarty & Million 1994). The optimal resection margin should not compromise local control from an inadequate resection or cause unnecessary functional morbidity from too much resection. There is no single definition for an adequate clinical resection margin (Byers et al 1978). This needs to be individualized with consideration of the patient and tumor related variables. The complexity of defining the adequate clinical margin is further compounded by the fact that certain variables, such as tumor thickness or the pattern of invasion at the tumor host interface may not be known until a final pathology report is rendered. Recurrence in the presence of a clear margin may result from many reasons such as the presence of residual tumor not identified by the pathologist (sampling error) or a margin of pre-conditioned epithelium or field change around the tumor.

The positive association of margin status with various tumor related factors indicates that negative margins are harder to achieve with large tumor loads, proximity to mandible and a posterior location in the oral cavity. In a similar fashion the ability to achieve a wide negative margin was associated with the use of composite resection and segmental mandibulectomy. Similarly, it would make sense that a wider margin would be necessary to achieve the goal of a histologically negative margin in a tumor that invades deeply as nests and cords of cells as opposed to a tumor with a broad and flat pushing margin (McMahon et al 2003). The presence of tumor at the resection margin is recognized as an indicator of more aggressive disease (Kademani et al 2005).

2.6.4 Post Removal Changes

The measured margins of oral cavity tumors shrink from when they are first removed in situ by the surgeon to the final measurement on the microscope slide by the pathologist. This shrinkage can range from 30 to 50%. Beaumont et al (1992) reported on the significant changes in the longitudinal diameter of whole SCC specimens from the in situ fresh states to the fixed state. The most significant alteration was in the measured diameter of the tumor with a mean shrinkage of 4.28 mm from the fresh to the fixed state. In this study there was a reduction of 46% from the planned surgical margin before resection to the time of microscopic clearance in the surgical pathology laboratory. Johnson et al (1997) reported a study on mongrel dog oral mucosa. They showed shrinkage rate in the labial buccal mucosa of 47.3% while the tongue shrinkage rate varied between 30.7% for the mucosal margin and 34.5% for the deep tongue margin. The greatest proportion of shrinkage occurred immediately upon resection (Johnson et al 1997).

2.6.5 Effect of Margin on Prognosis

Loree and Strong (1990) analyzed the significance of positive margins on survival in a large group of patients with a primary oral cavity carcinoma. A total of 398 patients were studied. The overall local recurrence rate in the entire positive margin group was twice the negative margin group, 36% versus 18% respectively. A positive margin also adversely affects survival. The five-year survival rate of the positive and negative margin groups were 52.7% and 60% respectively. In this study there was no improvement in overall survival or local control with the use of adjuvant radiotherapy.

In a study of 270 patients with tumors from a variety of oral cavity, pharyngeal and laryngeal sites Chen et al (1987) found the local recurrence rate in patients with negative margins to be 17%. For the 35 patients with positive margins in this study the local recurrence rate for microinvasive carcinoma, in situ, and close margin were 55%, 50% and 45% respectively. The proportion of patients surviving five years was significantly greater among those with negative margins.

In a recent study by Sutton et al (2003) from the Department of Maxillofacial Surgery in Liverpool, 200 consecutive patients undergoing primary surgery for previously untreated oral and oropharyngeal squamous cell carcinoma were evaluated. Of the 200 patients, 107 (53.5%) had a clear margin, 42% had a close margin and 4.5% had involved margin. They defined the clear margin as no evidence of tumor within 5 mm of the margin on histological assessment of formalin fixed tissue. They defined the close margin as tumor within 5 mm of the margin but with no evidence of tumor at the margin. They defined the involved margin as evidence of frank tumor at the margin. Their follow-up data showed that close surgical resection margins were associated with a shorter disease free and disease specific survival. At five years 78% of patients were alive and well or dead and free of disease, compared to 47% of those with close margins and 11% of those with involved margins. Patients with close surgical margins were at significantly greater risk of developing local recurrence (clear 12%, close 33% and involved 55%) or regional recurrence (clear 4%, close 10% and involved 22%). Their five year overall survival was 54% for patients with clear margin, 26% for patients with close margin and none of the involved margins survived five years. In their multivariate Cox regression analysis the relative risk of death associated with involved margin was 11.6, close margin was 2.66 and positive cervical nodes was 2.15.

McMahon et al (2003) reported their findings in two independent groups of patients. Two hundred and thirty-seven patients with oral and oropharyngeal carcinoma were treated in Sydney, Australia and 95 patients were treated in Lanarkshire, England. Using a univariate analysis a number of factors significantly predict the local recurrence in each group. In the Sydney group the condition of margins was one of them. While in the Lanarkshire group the status of the surgical margin did not predict the local recurrence with a univariate analysis. However, using a multivariate analysis only the observation of perineural invasion at the primary tumor site was an independent predictor of local recurrence. This applied to both groups of patients. When disease specific survival was studied on the univariate analysis the condition of margins was a significant prognostic indicator of the five-year disease specific survival in both groups. However, on multiple regression analysis only the presence and extent of regional nodal involvement independently influenced the disease specific survival. Their study concluded that for most patients who have a close or involved margin the biology of the disease influences the subsequent course irrespective of the width of clearance of tumor. This area of oral SCC margins requires further research and clarification.

2.7 Statistical Methods The use of simple methods of statistical analysis are routine in studies involving basic research. These methods include descriptive statistics such as the mean and its standard deviation for continuous variables, and percentage and sample size for categorical variables. Tests such as Chi-Square are also useful for categorical variables whereas t-tests are useful for continuous variables.

Kaplan-Meier curves comparing time-to success (survival) probabilities between the two study groups, using the Log-rank test, are more complex but aid in evaluating populations where survival from a disease is in question (Kaplan & Meier, 1958). A multivariate Cox’s proportional hazard model may be used to test the hypothesis of increased hazard or risk of death after controlling for potentially confounding variables, such as age and tumor stage (Cox, 1972). In addition, this model produces adjusted hazard ratios (HR) and the estimated adjusted probabilities of overall survival for each predictor in a particular model (Spruance et al 2004).

3 Aims of the study The purpose of this study is to expand the body of knowledge associated with the surgical treatment of oral squamous cell carcinoma. Therefore the specific aims of the study were:

1. To determine the treatment outcomes of squamous cell carcinoma of the retromolar trigone using a population-based study.

2. To determine the treatment outcomes of squamous cell carcinoma of the maxillary alveolus and

palate using a population-based study.

3. To describe the relationship of the maxillary artery and the lateral pterygoid muscle in a Caucasian sample and how this could impact on oral cancer care.

4. To observe the anatomy and function of the marginal mandibular branch of the facial nerve as it

pertains to the surgical management of oral cancer.

5. To test the significance of the positive surgical margin in oral cancer management.

4 Methods and Materials

4.1 Subjects This research project was performed involving a total of 715 subjects in whom oral cancer sites or related anatomy were the focuses of these studies. The work was divided into five studies. The numbers of subjects, with their demographic data, are listed in Table 1.

Table 1. Number, gender and mean age with range of the subjects participating in the five studies.

Study n (Male/Female) Age x (range) I (Patients) II (Patients) III (Cadavers) IV (Patients)

76 (56/20) 37 (12/25) 44 (21/23) 133 (78/48)

67.2 (56-78) 72.8 (59-85) 69.6 (58-74) 54.3(37-71)

V (Patients) 425 (250/175) 63.6(51-76)

n= number of male or female subjects. x=mean. The protocols to conduct these five studies and to review patient records were approved by the research

ethics committees and by the Department of Anatomy, University of Manitoba, Winnipeg, Canada, as part of a protocol entitled “A Quality Management Program for Head and Neck Oncology” (Protocol Reference Number: H2005:071) prior to the commencement of the studies.

4.2 Methods

4.2.1 Study I Retromolar Trigone Carcinoma

In study I, a total of 88 charts of patients registered with squamous cell carcinoma of the

retromolar trigone in the Province of Manitoba from January 1975 to January 2004 were evaluated. Twelve

patients were excluded from the study. This was because of incorrect site coding (n=3), insufficient clinical data (n=6), or pathology report absent or inconsistent with squamous cell carcinoma (n=3). Seventy-six cases of biopsy-proven previously untreated squamous cell carcinoma of the retromolar trigone were identified. One patient presented with bilateral disease 10 years apart. This patient was considered as two distinct cases.

Tumor staging was done using clinical data recorded at the time of initial assessment of each patient according to the TNM (tumor, node and metastasis) classification system of the International Union Against Cancer, 2002. After appropriate merges and data transformation, statistical analysis was done using SPSS software (SPSS Inc., Chicago, IL). Descriptive statistics were presented as mean ± standard deviation (SD) for continuous variables and percent (%) and sample size (n) for categorical variables. The comparison between the study groups was done for baseline characteristics using Pearson chi-square, with continuity correction when appropriate, or Fisher’s exact test for categorical variables and t-test or Mann-Whitney U-test for continuous variables. Kaplan-Meier curves comparing time-to success (survival) probabilities between the two study groups, using the Log-rank test, were used.

4.2.2 Study II Carcinoma of the Maxillary Alveolus and Hard Palate In study II, the records of 62 patients diagnosed with SCC of the upper jaw and registered in the Province of Manitoba from January 1975 to January 2004 were evaluated. Patients were excluded for the following reasons: incorrect site coding (n=6), insufficient clinical data (n=2), other pathology (n=16), and patients seen in consultation only (n=1). There were 37 cases of biopsy-proven and previously untreated SCC of the maxillary alveolus and hard palate (Figure 9). Staging was done using clinical data recorded at the time of the initial assessment of each patient according to the TNM (tumor, node, and metastasis) classification system of the International Union Against Cancer, 2002. After appropriate merges and data transformation, statistical analysis was performed using SPSS software (SPSS Inc., Chicago, IL). Descriptive statistics were presented as mean ± standard deviation (SD) for continuous variables, percentage (%), and sample size (n) for categorical variables. The comparison between the study groups was done according to baseline characteristics using the Pearson chi-square test with continuous correction when appropriate, Fisher’s exact test for categorical variables, and t-tests or Mann-Whitney U-test for continuous variables. Kaplan-Meier curves were used to compare time-to success or survival probabilities between the different study groups using the Log-rank test.

4.2.3 Study III Maxillary Artery

Study III is based on investigations in 44 cadavers (21 males and 23 females) that were embalmed and

preserved in 2 % formalin in the gross anatomy dissection laboratory at the University of Manitoba. All specimens were Caucasians of known age and sex.

The maxillary arteries were dissected bilaterally using a lateral infratemporal approach. The covering soft tissues including the skin, fascial layers, superficial portion of parotid gland, masseter and temporalis muscles were removed. The lateral surface of the ramus of the mandible was exposed along with the external carotid artery and the origin of the maxillary artery. This approach also exposed the mandibular condyles. Using an angled oscillating saw (Stryker, Kalamazoo, Michigan, USA), the zygomatic arches were removed and the mandibles were osteotomized at the neck of the condyles and at the symphysis menti. The attachment of the lateral pterygoid muscle to the pterygoid fovea and the tempromandibular joint was preserved. The mandible was then removed from its muscular attachments including the medial pterygoid, anterior belly of digastric, mylohyoid, geniohyoid, and genioglossus muscles without damaging any underlying tissues (Figure 10). These procedures were also carried out on the contra lateral side.

Careful dissection of the first and second parts of maxillary artery was performed so that the intimate relation of the second part of maxillary artery to the lower head of lateral pterygoid could be visualized. The superficial (Figure 10) or deep location (Figure 11) of the second part of the maxillary artery to the lower head of the lateral pterygoid muscle was recorded in every specimen. This relationship and any variations in symmetry were checked and documented. Photographs were taken using a digital camera (Nikon, Coolpix 4500, Tokyo, Japan). The information collected was statistically analyzed using the two- tailed binomial test assuming a 50% chance that the maxillary artery will be superficial or deep to the inferior head of the lateral pterygoid muscle.

4.2.4 Study IV Marginal Mandibular Nerve In study IV, clinical observation of the anatomy and function of the marginal mandibular nerve were

recorded in 133 neck dissections in 121patients. There were 78 males and 48 females with a mean age of 54.31±17.0 years. The neck dissections were done for squamous cell carcinoma (n=93) thyroid cancer (n=21), salivary gland malignancy (n=9), melanoma (n=8), angiosarcoma (n=1), and Levoise-Bensaud syndrome (n=1). The neck dissections were classified as comprehensive in 72 necks, selective in 59, and extended in 2. Level I nodes were removed in 109(83%) of the dissections. Twenty-five patients had received radiotherapy prior to the neck dissection.

All patients were positioned with a roll under the shoulders to maintain extension of the neck. Skin incisions in the upper neck were placed in skin creases 2 to 4 cm below the lower border of the mandible. In general skin flaps were elevated with sharp dissection in a sub-platysmal plane. When Level I was dissected an attempt was made to identify the marginal mandibular nerve using standard landmarks. This generally involved the observation of the nerve coursing superficial to the anterior facial vein. The distance from the lower border of the mandible to the lowest point of the nerve in the neck was measured. The relationship of the lowest point of the marginal mandibular nerve to the anterior facial vein and posterior facial vein (anterior division of retromandibular vein) was noted. These measurements and relationships were recorded with the head in a neutral position, and in an extended position with downward traction on the investing layer of the deep cervical fascia.

The functional status of the marginal mandibular nerve was recorded at the time of each postoperative visit. The patients were generally followed at intervals of 2 to 3 months. The average length of follow-up was 661.18 days (range: 7-1832 days; median: 541 days). Statistical analysis was performed with a Student's t-test, Chi-Square, or multivariate analysis where appropriate.

4.2.5 Study V Histologic Resection Margins In study V, an historical cohort of 707 patients with squamous cell carcinoma of the oral cavity from the

cancer registry of the Province of Manitoba, Canada from January 1975 to 2005 was examined. After excluding patients with incomplete records (n=34), seen in consultation only or treated with palliative intent (n=69), treated with radiotherapy as a single treatment modality (n=179) there were 425 cases of biopsy proven and previously untreated squamous cell carcinoma of the oral cavity managed with surgery with or without adjunctive radiotherapy. Pathology reports were reviewed for the status of the resection margin. The presence of tumor at the inked resection margin was considered as a positive margin. Tumor identified within 2 mm of the inked resection margin was recorded as a close margin. Tumors were staged using the TNM classification (International Union Against Cancer, 2002). Categorical data was evaluated using Pearson chi-square, with continuity correction when appropriate, or Fisher’s exact test. A t-test or Mann-Whitney U-test was used for continuous variables. For survival analysis Kaplan-Meier curves and log-rank test for comparing sub-groups was used A multivariate Cox’s proportional hazard model was used to test the hypothesis of increased hazard or risk of death with close and involved margins after controlling

for potentially confounding variables, such as age, tumor stage, etc., was done. In addition, this model produced the adjusted hazard ratios (HR) and the estimated adjusted probabilities of overall survival for each predictor in the model.

4.2.6 Statistics In studies I and II, descriptive statistics were presented as mean ± standard deviation (SD) for continuous variables and percent (%) and sample size (n) for categorical variables. The comparison between the study groups was done for baseline characteristics using Pearson chi-square, with continuity correction when appropriate, or Fisher’s exact test for categorical variables and t-test or Mann-Whitney U-test for continuous variables. Kaplan-Meier curves comparing time-to success (survival) probabilities between the two study groups, using the Log-rank test, were used.

In study III, the information collected was statistically analyzed using the two- tailed binomial test assuming a 50% chance that the maxillary artery will be superficial or deep to the inferior head of the lateral pterygoid muscle.

In study IV, the statistical analysis was performed using Student's t-test, Chi-Square, or multivariate analysis where appropriate.

In study V, categorical data was evaluated using Pearson Chi-Square, with continuity correction when appropriate, or Fisher’s exact test. A t-test or Mann-Whitney U-test was used for continuous variables. For survival analysis Kaplan-Meier curves and log-rank test for comparing sub-groups was used. A multivariate Cox’s proportional hazard model was used to test the hypothesis of increased hazard or risk of death with close and involved margins after controlling for potentially confounding variables, such as age and tumor stage, for example, was done. In addition, this model produced the adjusted hazard ratios (HR) and the estimated adjusted probabilities of overall survival for each predictor in the model.

5 Results

5.1 Retromolar Trigone Carcinoma There were 56 men and 20 women, with a mean age of 67.22±10.3 years. Tobacco or alcohol use, or a combination of both was documented in 66 (86%) patients (Table 2). Table 2. Demographics of patients with squamous cell carcinoma of retromolar trigone.

Parameter Number

Male 56 (74%)

Female 20 (26%)

Age at presentation 67.26 +/-10.3 years

Tobacco and / or alcohol use 66 (86%)

Diabetes, hypertension and ischemic heart disease were the most common comorbidities recorded. Twenty-three patients had 34 second primary tumors, of which 62% were synchronous and 55% involved the upper aerodigestive tract. Eight patients had Stage I lesions, while 25, 12, and 29 patients had Stage II, III, and IV disease respectively (Table 3). Two patients did not undergo staging. Twenty-eight patients (36%) had clinically positive nodes at the time of presentation.

Table 3. Stage of squamous cell carcinoma of retromolar trigone at diagnosis and survival.

Stage of Disease Number of Patients Five Year Survival

At Diagnosis

I 8 75%

II 25 57%

III 12 56%

IV 29 38%

Twelve patients were treated with a palliative intent. Surgical treatment was used as the primary modality

in 20 (26%) patients. Fifteen patients received postoperative radiotherapy. Radiotherapy was used with curative intent in 29 patients (39.5%). Three of these patients had a neck dissection for persistent disease. Intraoral excision of the primary tumor was performed in 5 patients, while the reminder required a cheek flap for exposure. Thirty-two patients underwent mandibular resection, either a marginal resection (n=10) or segmental mandibulectomy (n=22). Primary closure was possible in 11 patients. Margins on permanent section were clear in 18 patients, close (within 2 mm) in 8 patients, and involved in 7 patients. Radiation therapy as a primary treatment modality or as adjunctive treatment was delivered by conventional fractionation at a dose ranging from 4350 to 6600 cGy. Twenty-nine patients were managed with neck dissection alone. Ten patients were treated with neck dissection and adjunctive radiotherapy. Thirty-seven patients were treated with radiation therapy alone. Twenty-eight patients failed treatment, 18 (64%) at the primary site, 3 (10.7%) in the neck, 2 (7%) at the primary site and in the neck. Four patients had distant metastasis, and the site of failure was not documented in one patient. The overall median survival time was greater than 60 months and 51.4% of the patients survived the 5-years period. When disease specific survival was considered 67.7% of the patients survived the 5-years period and median survival time was still above 60 months. The median disease free survival was 60 months and 40.3% of the patients survived 5 years. There was no statistically significant difference between stage of disease and overall survival. Five-year overall survival rates were 75%, 57%, 56%, and 38% for Stage I, II, III and IV respectively, (p=0.2230) (Table 3). Five-year overall survival rates for patients with N0 and N+ neck disease were 55% and 47.5% respectively (p=0.5712). The 5-year overall survival by treatment modality for patients treated with surgery alone was 64%, surgery and radiotherapy 57%, and 45% when radiation was elected as a single treatment modality (p=0.1260). The 5-year overall survival for patients with clear, close (within 2 mm) or involved surgical margins was 68%, 83%, and 0.0%, respectively, (p=0.0270).

5.2 Carcinoma of the hard palate and maxillary alveolus

The tumor epicenter was found to involve the maxillary alveolar ridge in 26 patients and the hard palate in the remaining 11 patients (Figure 9). There were 12 males and 25 females, with a mean age of 72.8 years (SD12.1). Tobacco, alcohol consumption, or a combination of both was documented in 19 (50%) patients. One patient was a reverse cigarette smoker. Diabetes, hypertension, and ischemic heart disease were the most common comorbidities recorded. Eight of the patients had a total of 11second primary tumors, of which 42% were synchronous and involved the upper aerodigestive tract. Fourteen patients (38%) had pain as the major presenting symptom. Six patients had stage I lesions, 9 stage II, 4 stage III, and 15 patients stage IV. Three patients could not be staged. Seven patients (19%) had clinically positive nodes at the time of presentation (Table 4). Table 4. Stage of disease at time of presentation. Three patients were not staged.

Stage of disease Number of patients

Stage I

Stage II

6

9

Stage III 4

Stage IV 15

Figure 9. Diagram representing the location of tumors in the 37 patients with oral squamous cell carcinoma of the maxillary alveolus and hard palate.

Eight patients (21.6%) were treated with a solely palliative intent. Surgery was used as the primary treatment modality in 23 patients with 9 of these patients receiving postoperative radiotherapy. Radiation therapy was used as a single treatment modality in five patients (13.5%). One patient was treated with radiotherapy and chemotherapy (Table 5).

Table 5. Treatment by stage.

Stage at

Diagnosis Surgery Radiotherapy Combination

Stage I 5

Stage II 7 1 1

Stage III 1 3

Stage IV 2 1 6*

Not Staged 2

(Eight patients were treated with palliative intent. * One patient was treated with radiation therapy and chemotherapy.)

In the 23 patients treated with surgery a transoral excision was used in 11 cases. In eleven patients a cheek flap was deemed necessary for adequate exposure. An operative report was not identified in one patient. Margins on permanent section were found to be clear in 16 patients, close or within 2 mm in 3 of the patients, and involved in 4 patients. Radiation therapy, whether used as an adjunctive or as a primary treatment modality, was delivered by conventional fractionation at doses ranging from 4,125 to 7,000 CGy. Four of the seven patients presenting with positive lymph nodes were treated with a curative intent. Two of these patients were treated with radiation alone. In the remaining 25 node negative patients who were treated with curative intent, the neck was treated electively in 2. Ten of the 29 patients treated with curative intent failed treatment with 6 failing at the primary site and 4 failing in the neck. The absolute and disease free survival at 5-years was 33% and 62% respectively (Figures 3 and 4). The 5-year disease-free survival (Figure 5) was 82% for stage I & II and 48% for stage III and IV (p=0.056). The initial treatment modality whether surgery, radiotherapy, or a combination of both significantly influenced the 5-year survival rate (p=0.001). No patient treated with radiotherapy as a single treatment modality survived 5 years or longer. The 5-year disease- free survival for patients treated with surgery alone, and a combination of surgery and radiotherapy was 69% and 73% respectively. Margin status did not affect survival in any of the cases with involved margins.

5.3 Maxillary Artery

Superficial (lateral) course of the maxillary artery to the lower head of the lateral pterygoid muscle

The lateral or superficial position of the second part of the maxillary artery in relation to the lower

head of the lateral pterygoid muscle was found in 68% of all the specimens examined. In the male cadavers, the second part of the maxillary artery was found to be superficial in 71% of the total number of cases (p= 0.052). On the other hand, in the female cadavers, the artery was found to be superficial to the lower head of the lateral pterygoid muscle in 65% of the total number of cases (p= 0.117). There was no statistically significant variation in the course of the maxillary artery between both sides of the same cadaver and between the sexes (Figure 10).

Figure 10. Superficial course of the maxillary artery over the lower head of the lateral pterygoid muscle. Lateral pterygoid muscle (LPTM). Maxillary artery (MA).

Deep (medial) course of the maxillary artery to the lower head of lateral pterygoid muscle

The frequency of the deep (medial) course of the second part of maxillary artery was relatively

low. The maxillary artery was found to run deep to the lower head of the lateral pterygoid muscle in only 32% of all the cases. In the male cadavers, the incidence was 29%, while in the female cadavers it was 35% (Figure 11). While there was a trend for female cadavers to display the deep course of the maxillary artery more often than males, this difference was not statistically significant. Similarly, there was no statistically significant variation in the deep course of maxillary artery between the two sides in the same specimen and between both sexes.

Figure 11. Deep course of the maxillary artery to the lower head of the lateral pterygoid muscle. Lateral pterygoid muscle (LPTM). Maxillary artery (MA).

5.4 Marginal Mandibular Nerve The marginal mandibular nerve was managed with the intent to preserve function in 112 (84%) necks. The nerve was visualized in 76 (68%) of these patients. A decision to sacrifice the nerve, based on oncological considerations, was made in 21 dissections (Table 6).

Table 6. Management of the marginal mandibular nerve in 133 neck dissections. (Numbers in parenthesis are percentages)

Management Intent Nerve Visualized Nerve Not Visualized Total

Preserve

76(68) 36(32) 112

Sacrifice 9(43) 12(57) 21

Total 85(64) 48(36) 133

Observations on the course and surgical relations of the marginal mandibular nerve were recorded in 85 patients. With the head in neutral position the nerve was at its lowest point 0.31±0.4cm (range 0.0-1.5 cm) below the inferior border of the mandible in the region of the posterior facial vein. (Figure 12) With the neck extended the nerve was displaced in an anterior and downward direction with the lowest point 1.25±0.7cm (range 0-3.0cm) below the mandible between the posterior and anterior facial veins (p<0.01). In the extended position the nerve was >1 cm below the lower border of the mandible in 42(54%) and >2 cm in 8(10%) of patients. (Table 7).

Table 7. Distance marginal mandibular nerve identified below the mandible with the neck in neutral and extended position in 78 neck dissections. (Numbers in parenthesis are percentages).

Distance Below Mandible Neutral n=78

Extended n=78

>1 cm 3(4) 42(54)

>2 cm 0(0) 8(10)

Figure 12. The relationship of the marginal mandibular nerve (MMN) to the lower border of the mandible (MLB), anterior facial vein (AFV) and posterior facial vein (PFV) with the neck in neutral position (A) and in extended position (B) with downward traction on the cervical fascia.

The functional status of the nerve was recorded following 121 neck dissections. Neuropraxia was recorded in 18(90%) of patients where the nerve was visualized and deliberately divided (n=9) or presumed divided with dissection in a plane superficial to platysma (n=11) for oncologic reasons. In the remaining 101 patients where the intent was to preserve the nerve praxia was observed in the immediate postoperative period in 29 and persisted in 16(16%). Temporary dysfunction of the marginal mandibular nerve had usually resolved by the second or third postoperative visits (3 to 6 months). Neuropraxia was not observed in the 24 patients where level I lymph nodes were not dissected. The incidence in the 77 patients where a level I dissection was performed with the intent to preserve nerve function was 16/77(21%). The relationship between nerve dysfunction and clinical and treatment related factors in this subset is shown in Table 8. The incidence of nerve dysfunction appeared to be higher when the neck dissection followed radiotherapy. This approached, but did not achieve statistical significance. Visualization of the nerve during neck dissection did not impact the functional outcome. The use of multivariate models did not establish any independent predictors of nerve dysfunction.

5.5 Resection Margin Status The mean age of the cohort was 63.6±12.6 years with 58.89% males and 41.11% females. Tobacco or alcohol use or a combination of both was documented in 80% of patients. One hundred and ninety-eight -second primary malignancies were identified in 145 patients, of which 61% were metachranous. The tongue was the primary site in 129 (30.4%) of patients, floor of mouth in 169 (39.8%), lower alveolus in 38 (8.9%), buccal mucosa in 33 (7.8%), upper alveolus in 23(5.4%), and retromolar trigone in 33 (7.8%). One hundred and twenty-four patients had Stage I lesions, while 118, 68, and 105 patients had Stage II, III, and IV disease respectively. Ten patients could not be staged. One hundred and seven patients (26%) had clinically positive nodes at the time of presentation. Table 8. Relationship of the status of the surgical margin to treatment parameters.

Variable Involved (n=62) Close (n=68) Clear (n=295)

Resection

Per-oral(n=243) 31(12.8%) 40(16.5%) 172(70.7%)

Composite(n=182) 31(17.1%) 28(15.3%) 123(67.6%)

Mandibulectomy

Marginal(n=62) 8(12.9%) 14(22.6%) 40(64.5%)

Segmental(n=113) 23(20.3%) 15(13.3%) 75(66.4%)

None =(n=250) 31(12.4%) 39(15.6%) 180(72%)

Frozen Section

Yes(n=225) 36(16.0%) 31(13.8%) 158(70.2%)

No(n=200) 26(13.0%) 37(18.5%) 137(68.5%)

Adjunctive Radiotherapy

Surgery(n=298) 31(10.4%) 47(15.8%) 220(73.8%)

Surgery + Radiotherapy(n=127) 31(24.4%) 21(16.5%) 75(59.1%)

Surgery was used as a single treatment modality in 298 (70.1%) patients and 127(29.9%) received

adjuvant radiotherapy (Table 8). Per-oral excision of the primary tumor was performed in 243 patients and 182 were managed with a composite resection. Mandible was resected in 175 patients, a marginal resection in 62 and segmental in 113. The clinically negative neck was treated electively in 120 patients (39.0%). Adjunctive radiotherapy was delivered by conventional fractionation, to a median dose of 5000 CGy (mode 5000 CGy). There were 4 deaths within 30 days of treatment completion. Minor morbidity was recorded in 23 patients treated with surgery (7.7%) and 42(33.1%) treated with adjunctive radiotherapy (p<0.0001). Major morbidity was observed in 8 (2.7%) of patients treated with surgery and in 10 (7.9%) of patients receiving radiotherapy (P=0.0313).

Sixty-two patients (14.6%) had microscopic tumor identified at the inked resection margin (Figure 8), 68 (16.0%) patients had close (<2mm), and 295 (69.4%) patients had clear (>2mm margins). The relationships between the status of the surgical margins and clinical data is shown in Table 8. Involved margins were observed less frequently following excision of tongue tumors, however the observed differences did not achieve statistical significance (P=0.1239). Involved margins were associated with both advanced T stage (p=0.0033) and the presence of clinically positive lymph nodes (p=0.0034). The type of surgical resection did not influence the incidence of involved margins (Table 8). The use of intra-operative frozen section was not associated with either involved or clear margins. Adjunctive radiotherapy was used more frequently in patients with involved margins (p=0.0005).

Table 9. Relationship of the status of the surgical margin to the incidence of recurrence, sites of initial recurrence, and 5-year overall survival (numbers in parenthesis are percentages).

Involved(n=62) Close(n=68) Clear(n=295) All(n=425) P-Value

Recurrence 25(40.3) 26(38.2) 73(24.7) 124(29.2) P=0.009

Site of Recurrence

Primary Site 17(27.4) 16(23.5) 41(13.9) 74(17.4) P=0.005

Neck 14(22.6) 11(16.2) 37(12.5) 62(14.6) P=0.116

Distant Metastases 3(4.8) 5(7.4) 12(4.1) 20(4.7) P=0.207

Overall Survival at 5 years

38.7% 58.3% 68.4% 61.9% P<0.0001

Recurrent disease, observed most frequently at the primary site and neck, was observed in 78(43%) of patients treated with surgery and 46(36%) of patients treated with adjunctive radiotherapy (P=0.24). The relationships between margin status and the sites and incidence of recurrent disease and overall survival is shown in Table 9.

Table 10. Relationship of the status of the surgical margin to the incidence of recurrence, sites of initial recurrence, and 5-year overall survival (numbers in parenthesis are percentages).

Involved(n=62) Close(n=68) Clear(n=295) All(n=425) P-Value

Recurrence 25(40.3) 26(38.2) 73(24.7) 124(29.2) P=0.009

Site of Recurrence

Primary Site 17(27.4) 16(23.5) 41(13.9) 74(17.4) P=0.005

Neck 14(22.6) 11(16.2) 37(12.5) 62(14.6) P=0.116

Distant Metastases 3(4.8) 5(7.4) 12(4.1) 20(4.7) P=0.207

Overall Survival at 5 years

38.7% 58.3% 68.4% 61.9% P<0.0001

Patients with involved and close margins had a similar incidence of treatment failure, which was significantly higher than that observed in patients with clear margins (Table 10). Failure at the primary site was more frequent in the presence of both involved and close margins. Margin status was not associated with failure in the neck or distant metastases. One hundred and ninety-eight deaths were documented, 73 attributed to oral cancer, 51 to second primaries, 44 from other conditions, and could not be accurately determined in 30. Sixty-two percent of the cohort survived 5 years. Overall survival was significantly worse for patients with involved margins when compared to those with close or clear margins (Figure 1). The Cox’s proportional hazard model is shown in Table 11. Age, T Stage, N Stage, and involved margins had a significant impact on 5-year survival. After controlling for the confounding variables the presence of tumor at the inked resection margin increased the risk of death by 90% (HR 1.9). The presence of a close margin did not have a significant impact on survival.

Table 11. Cox’s proportional hazard model predicting 5-year overall survival.

Variable Hazard Ratio

95% CI P-Value

Included in the model

Involved vs Clear/Close Margin 1.9 (1.2,2.9) 0.0026

Age ≥65 vs Age <65 1.5 (1.1,2.2) 0.0017

T Status 1.4 (1.2,1.6) 0.0002

N Status 1.3 (1.0, 1.6) 0.0465

Excluded from the model

Close vs Clear Margin 0.1027

Gender 0.3212

Treatment Modality 0.6105

Stage 0.8805

Site 0.8379

6 Discussion

6.1 Retromolar Trigone Carcinoma This series of 76 patients with RMT squamous cell carcinoma is comparable in many respects, including the age and sex distribution, etiologic factors, and incidence of second primaries, to most series of oral cancer. However, in contrast to other series this is a population-based study, describing events over a decade within a well-defined geographic boundary, the Province of Manitoba in Canada, with a population of 1.5 million. The authors have developed the clinical impression that patients with early stage disease of the RMT fail treatment more frequently than observed with similar staged lesions of the tongue and floor of mouth. Survival of greater than 80% has been reported for early staged lesions of the tongue and floor of mouth (Lehman et al 1982, Shaha et al 1984, Franceschi et al 1993, Hicks et al 1997, Woolgar et al. 1999). The previously reported prognosis for squamous cell carcinoma of the RMT is relatively poor with overall survival ranging from 26% to 80% (Byers et al 1984, Lo et al 1987, Kowalski et al 1993, Huang et al 2001, Genden et al 2003, Ayad et al 2005, Mendenhall et al 2005). Unlike the previous studies the overall survival of 51.4% in our series does not reflect a selection bias for patients and initial treatment modality such as radical surgery (Kowalski et al 1993) or treatment with radiation alone (Lo et al 1987, Huang et al 2001, Ayad et al 2005). The outcome reported in this study includes all patients managed specifically with squamous cell carcinoma of the RMT. Clinical stage and survival did not show a significant correlation in this review. The lack of significance may be a function of the small sample number in each of the subdivided groups. While Kowalski’s team reported similar findings (Kowalski et al 1993); other investigators (Ayad et al 2005, Mendenhall et al 2005) have shown a significant influence on stage and survival with both univariate and multivariate analysis. The initial treatment modality did not affect survival in this series, which is again consistent with previous reports (Byers et al 1984, Lo et al 1987, Kowalski et al 1993). Surgery seems to have the best survival rate, even if this is not statistically significant. The lack of success may have been due to the small sample number in each treated group. More recent reports of treatment outcomes for squamous cell carcinoma of the RMT suggest that survival is improved with surgery followed by adjunctive radiation therapy (Huang et al 2001, Antoniadoes et al 2003, Mendenhall et al 2005). The surgical margin status impacted local-regional treatment failure as reflected by a significant decrease in overall survival with positive margins. This was the only significant predictor of outcome in this series. This is consistent with some previous reports involving other sites and types of oral cancer (Looser et al 1978, Batsakis JG 1999). However the margin status was not found to predict survival in some other series (Kowalski et al 1993, Huang et al 2001). This outcome was attributed to the use of postoperative radiotherapy by one investigator (Huang et al 2001). The lack of concordance between studies with respect to prognostic factors for squamous cell carcinoma of the RMT again reflects small sample sizes with differences in patient selection. Larger multicenter population based studies using appropriate multivariate models are necessary to resolve these

issues. There has been no agreement on the optimal management of these tumors of the RMT. Early reports recommended surgical treatment over radiotherapy (Barbosa 1959) with reports of experiences with extended commando operations in 114 patients with or without postoperative radiation resulting in a 5-year absolute survival rate of 55.3% (Kowalski et al1993). The conclusion in this series was that adjunctive radiotherapy improved local and regional control (Kowalski et al 1993). The MD Anderson experience showed a recurrence rate of 16% in patients treated with radiation therapy alone and 18% in patients treated with post operative radiation (Byers et al 1984). The 5 year absolute survival was 26%. In another report from the same institution the experience with irradiation of carcinoma of the anterior faucial pillar and the RMT showed the 5-year determinant survival to be 83% (Lo et al 1987). A more recent study reported a 5-year actuarial survival of 60.6% in 50 patients treated with surgery with 24 receiving adjuvant therapy (Hao et al 2006). Another review of 65 patients showed a 5-year disease free survival of 31% with radiation alone, compared to 90% and 63% with surgery and pre- or postoperative radiation respectively (Huang et al 2001). These authors recommended combined surgery and radiotherapy for all stages of disease. The most recent review in 2005 reported 5-year overall survivals of 40% and 56% in 35 patients managed with radiation alone and 64 managed with surgery and radiotherapy respectively (Mendenhall et al 2005). Using a multivariate analysis this study found the stage (p=0.0476) and the treatment modality (p=0.0098) to be significant predictors of survival (Mendenhall et al 2005). One other study has shown improved survival with the addition of chemotherapy to surgery and radiation (Lo et al 1987). The optimal management of squamous cell carcinoma of the RMT cannot be determined from this or any other review. When considering relatively uncommon tumors such as squamous cell carcinoma of the RMT there is a need for the head and neck oncology community to pool population based outcome data to establish accurate prognostic indicators.

6.2 Carcinoma of the hard palate and maxillary alveolus Squamous cell carcinoma of the maxillary alveolus and hard palate is relatively uncommon in western society (Bhansali 1961, Ong 1970, Kroll & Hoffman 1976). It has been noted that neoplasias, both SCC and non SCC, of the upper jaw constitute only 2% of all head and neck malignancies, and 10% of oral cancers (Petruzzelli & Meyers 1994). These lesions appear to be considerably more common in India accounting for 40 to 55% of the oropharyngeal cancers (Ramulu & Reddy 1972, Reddy 1974). The relatively low numbers of these tumors is the most likely reason these lesions are often grouped and reported together with other sites such as the mandibular alveolus (Cady & Catlin 1969, Soo et al 1988, Gomez et al 2000) and soft palate (Cady & Catlin 1969, Evan & Shah J 1981, Shibuya et al 1984) or combined with salivary gland tumors (Evan & Shah 1981). In contrast to most other reports this study examines treatment outcomes in a site which is anatomically specific and is pathologically limited to squamous cell carcinoma of the maxillary alveolus and hard palate in a population based historical cohort. The anatomical characteristics of the maxillary alveolar ridge and hard palate are similar as together they form the roof of the oral cavity and the floor of the nasal cavity and maxillary sinus. Full thickness resection of the structures involved by tumor at this site will result in oro-nasal and oral antral communication, a problem unique to this site. This unique anatomic relationship makes the radiographic work-up with CT scans and MRI scans crucial to determine tumor extent and stage. This radiographic visualization ultimately helps decides the staging, the extent of surgical resection and the use of other modalities of treatment. Therefore, the clinical presentation, the surgical management and reconstructive considerations including the management of oro-nasal and oral antral communication with immediate obturation, are similar for these sites (Figures 1a to 1d). From a clinical perspective it is often difficult in advanced tumors to determine the exact epicenter (Figure 2). The maxillary alveolus and hard palate, in contrast to the remainder of the oral cavity, seems to share a female predilection in this and other reports (Chung et al 1979, Chung et al 1980, Yoruzu et al 2001). In searching the literature further, the authors were unable to find reasons for this female predilection as of yet. However, based on the foregoing, the authors feel that it is most appropriate to consider SCC of the upper alveolus and hard palate as a single entity.

The overall or absolute survival of 33% at 5 years reported in this study is relatively low. Reported survival rates for these sites which range from 21 to 76% (Cady & Catlin 1969, Evan & Shah 1981) are difficult to interpret as they are often grouped to include other sites and other pathological entities such as salivary gland tumors. The Improved disease free survival at 5 years of 62% reported in the current study reflects the influence of advanced age of the patients with significant co-morbidities on overall survival. These considerations as well as the frequency of stage IV disease (40%), account for the relatively large number of patients not amenable to treatment with curative intent. Margin status did not affect survival or outcome in any of the cases with involved margins. The reality is that this is a study of only 37 cases in a rare site and the author’s did not feel comfortable drawing conclusions from such a small sample. The number of participants in this study is too small to examine prognostic factors in detail. In this report there was a trend towards decreased survival with advanced stage of disease as observed in other series (Chung et al 1979, Yoruzu et al 2001). Yokoo et al (2002) were unable to correlate stage of disease with prognosis. In their series 80% of patients had stage IV tumors. They proposed a new classification system dependent upon the involvement of the maxillary sinus or nasal floor. The authors of the present study observed improved survival in patients treated with surgery. Arguments in favor of one treatment modality over another in this, and the other series reviewed, are of limited value because of small sample size in the studies. There seems to be better prognosis for Stage III and IV lesions at this site compared to the late stages at other oropharyngeal sites. This trend was confirmed in other series (Chung et al 1979, Yoruzu et al 2001, Yokoo et al 2002). The reasons for this are unknown and not apparent in the literature. The authors feel that this may be due to differences in lymphatic drainage at the maxillary alveolar and palatal sites when compared to other oropharyngeal sites. This tumor is more prevalent in India and these numbers may also represent some geographic differences potentially due to differences in social habits such as reverse cigarette smoking. The authors feel that the numbers in the current study are too small to be able to make further conclusions at this time. In conclusion squamous cell carcinoma of the maxillary alveolar ridge and hard palate differs from other oral cancers in several aspects. SCC tends to occur at these sites in a relatively older population with a female predilection. A relatively high number of patients were not amenable to treatment with curative intent. Surgical treatment with or without radiation therapy, appears to improve disease control. A future prospective multi-center study with a larger cohort of patients may help to answer questions regarding the prognosis of treatment related to the specific stages of this disease.

6.3 Maxillary Artery Understanding the course of maxillary artery in the infratemporal fossa with its variation is of potential of interest to many practitioners who treat patients with lesions of the cranio-maxillofacial region. For instance, tumors arising or extending into the infratemporal fossa via foramen ovale or foramen rotundum, such as juvenile nasopharyngeal angiofibromas, are commonly excised by lateral cranial base approaches such as the Fisch type C (Allen et al 1974, Fisch & Pilsbury 1997) or subtemporal / infratemporal approaches (Fisch 1978, Fisch 1982, Gates 1988, Holliday 1986, Mickey et al 1988). In these approaches, the entire contents of the infratemporal fossa are exposed including the maxillary artery, which may be greatly susceptible to damage or sacrifice, depending on the extent of the surgery and the size of the tumor. Therefore, knowing the course of the maxillary artery and its variations will aid in surgical planning and in minimizing intraoperative or postoperative hemorrhage.

In orthognathic and temporomandibular joint surgery, many surgical procedures are performed in close proximity to the maxillary artery in the infratemporal fossa. Mandibular osteotomies, such as intra-oral vertical ramus osteotomy (IVRO) or bilateral sagittal split osteotomy (BSSO), are commonly performed to correct dentofacial deformities (Fujimura et al 2006). Open temporomandibular joint procedures using the preauricular approach are relatively commonly used. These surgical procedures or approaches have been developed to improve safety in order to minimize or prevent damage to vital anatomic structures such as

the maxillary artery. Avoidance of the maxillary artery or to one of its branches is necessary to prevent extensive hemorrhage intraoperatively or postoperatively (Apinhsmit et al 2004, Fujimura et al 2006).

In clinical dentistry, mandibular blocks (Coleman & Smith 1982) such as Gow-Gates or Akinosi, are very commonly used techniques to deliver local anesthetic agents prior to many dental procedures. The site of injection is extremely close to the branches of the maxillary artery, which may result in undesirable consequences, such as intrarterial injection or hematoma (Frangiskos et al 2003). The clinician must be fully aware of the course of the maxillary artery and its variations in order to prevent the side effects from intra-arterial local anesthetic injection.

The course of the maxillary artery in relation to the inferior head of the lateral pterygoid muscle has been a controversial point. As a result, a number of studies were conducted to investigate this mutual relationship. A review of the literature showed that many authors found the maxillary artery superficial to the lower head of lateral pterygoid muscle. This finding ranged between 54% and 70% (Thompson 1891, Lurje 1946, Laskar et al 1951, Skopakoff 1968, Vironis 1996, Orbay 2006). Moreover, Adachi (1928) found a higher frequency of superficiality (93%) in the Japanese population, raising the possibility of racial differences. However, Lauber (1901) in his investigation showed only 8.5% of the Caucasian population with a superficial maxillary artery. In the present study, the relationship of maxillary artery to the lower head of the lateral pterygoid muscle was generally similar to that of other studies. On the one hand a trend was noted in the lateral course of the maxillary artery relative to the lower head of the lateral pterygoid muscle to be more common in males (71%) than in females (35%), supporting other data in the literature (Thompson 1891, Lurje 1946, Laskar 1951, Skopakoff 1968, Vrionis et al 1996, Orbay et al 2006). On the other hand, the medial course was less common in males (29%) than in females (35%) in the present study. However, the course of the maxillary artery in relation to the lower head of the lateral pterygoid was symmetrical in all of our specimens. This is not in agreement with the report by Pretterklieber and co-workers, which is the only study that shows asymmetry to exist in the course of the maxillary artery between the two sides in the same specimen (Pretterklieber et al 1991). This finding is questioned because there is no other supporting work in the literature.

Many authors attempted to explain the potential factors that determine the position of maxillary artery relative to the lateral pterygoid muscle in the infratemporal fossa. However, craniometric analysis, including facial height, cranial breadth or length, and cephalic index, did not reveal any correlation between the variable position of the maxillary artery and the measured craniometric parameters (Skopakoff 1968, Macho 1986, Sjovold 1988, Vertcauteeren 1990, Pretterklieber et al 1991, Vrionis et al 1996, Orbay et al 2006). In another study the lateral course of the maxillary artery was found to be more common in brachycephalic and mesocephalic individuals (Lurje 1946). However, this study by lacked any statistical evaluation and was contradicted by other studies (Laskar et al 1951). One theory, the metabolic hypothesis (Adair et al 1990) was proposed to explain the potential factors involved in angiogenesis and the vascularization of growing tissues, however, it failed to explain the factors responsible for variation in the position of the second part of the maxillary artery relative to the lateral pterygoid muscle.

In conclusion, this study confirms some of the earlier findings concerning the more common superficial position of the maxillary artery in relation to the lateral pterygoid muscle. However, the findings that there was a trend for the deeper course of the second part of the maxillary artery to be slightly more common in females than in males, and that there was no evidence of asymmetry in the course of the artery within the two sides of the same specimen, were unique to this study. The mechanism responsible for determining the relative position of the artery to the lateral pterygoid muscle is still unknown. This study helps to begin to define the question of the effect of ethnicity on the relative position of the maxillary artery to the inferior head of the lateral pterygoid muscle by studying a Caucasian population from Canada. Comparison with future studies in other ethnic groups (Apinhsmit et al 2004) will further help to answer this question.

The relationship between the maxillary artery and the lateral pterygoid muscle may be important during major facial reconstruction, cleft craniofacial reconstruction, cancer surgery, and surgery to the temporomandibular joint. The anatomy of the terminal end of the maxillary artery should be kept in mind during surgical dissection.

6.4 Marginal Mandibular Nerve The complication of marginal mandibular nerve injury following neck dissections for malignant disease has not received much attention when compared to other complications of head and neck surgery (Moffat & Ramsden 1977). This is not surprising as historically other morbidities of major head and neck resections have over shadowed this disability as they were much more overwhelming. The significance of asymmetry of the lower lip increases as morbidity from other aspects of the treatment is minimized through multimodal treatment plans that incorporate conservative surgical approaches and organ preservation protocols. The frequency of nerve dysfunction observed in this prospective evaluation of neck dissections, is comparable to that reported in retrospective reviews of outcome following excision of the submandibular gland for benign disease (Hald & Andreassen UK 1994). It is recognized that retrospective studies underestimate the true incidence of complications (Hald & Andreassen 1994, Tulley et al 2000). This series represents a heterogeneous population of patients with respect to pathology, clinical, and treatment related parameters including the extent of neck dissection. In the cases where praxia of the marginal mandibular nerve was observed following a neck dissection with visualized or presumed division of the nerve, the outcome of the nerve could usually be predicted preoperatively. These patients generally had advanced local or regional disease. Patients where level I nodes were not dissected were included in the functional assessment as the proximal marginal mandibular nerve is exposed to potential traction type injury from retraction to expose high jugulodigastric lymph nodes. The absence of nerve dysfunction in this cohort of patients may be explained by the fact that the nerve is well protected by surrounding soft tissue. When level I was dissected and the intent was to preserve the nerve the incidence of praxia appeared to be higher after preoperative radiotherapy. The majority of patients in this series were managed surgically within 6 to 8 weeks of completion of radiotherapy. Tissue planes tended to be edematous, however the anatomy was not distorted. The fact that no independent predictors of nerve dysfunction were identified in patients where level I was dissected suggests that anatomic considerations are paramount in preserving the function of the marginal mandibular nerve. The anatomy of the marginal mandibular nerve in its vulnerable position in the upper neck is variable between patients. The course and position of the nerve also varies with the position of the head and downward traction on the investing layer of cervical fascia. The dynamic nature of the nerve was evident in this study. This explains, in part, the discrepancies between cadaver studies and clinical observations. Cadaver tissue is contracted and relatively immobile. Of interest when fresh, as opposed to embalmed cadaver material is studied, the nerve is consistently reported below the lower border of the mandible. In a series of 22 fresh cadavers found the nerve below the mandible in 63% of specimens posterior to the facial artery and in 27% anterior to the artery (Savary et al 1997). Savary et a (1997) observed some branches as low as 3-4 cm below the lower border of the mandible. The marginal mandibular nerve was well below the inferior border of the mandible in almost every instance in fresh cadaver specimens and clinical dissections (Nelson & Gingrass 1979). These authors therefore recommended cervical incisions several centimeters below the inferior border of the mandible (Nelson & Gingrass 1979). The present study confirms the consistent course of the marginal nerve below the inferior border of the mandible. The observation that the nerve is drawn downward when the neck is extended (Baker & Conley 1979) has been confirmed by the present study. The placement of cervical incisions 2 cm below the lower border of mandible will place the nerve at risk in a significant number of patients (Baker & Conley 1979). Detailed anatomic dissections (Dingman & Grabb 1962, Nelson & Gingrass 1979, Skandalakis et al 1979) show that there are often multiple contributions to the marginal mandibular nerve and that it can be difficult to discern between mandibular and platysmal branches. Dingman and Grabb noted that in many specimens fine branches could be seen running along the lower border of the mandible, some up to 2 cm below it (Dingman & Grabb 1962). They stated that these branches terminated in and innervated the platysma in all specimens. Nelson and Gingrass argued that these same branches, when followed anteriorly in fresh cadaver specimens and in clinical practice, ascended over the lower border of the mandible to innervate specific lip depressors. They contend that these branches should be identified as mandibular and not cervical branches (Nelson & Gingrass 1979). Skandalakis and co-workers described an anterior ramus of the cervical division of the facial nerve that joins the mandibular ramus to contribute to the innervation of the lower lip (Skandalakis et al 1979). There are also observations that the platysma, innervated by the cervical division of the facial nerve, contributes to depression of the lower lip (Tulley et al 2000). Division of the cervical branch of the facial nerve or the platysma can result in a pseudo-paralysis of the marginal

mandibular nerve that usually recovers spontaneously (Ziarah & Atkinson 1981). Detailed dissections of the marginal mandibular nerve were not done in our study. From a practical point of view there is usually a single identifiable nerve rami coursing superficial to the facial veins which results in asymmetry of the lower lip if divided. Stern indicates that if the structure in question during dissection does not course above the inferior border of the mandible within 2 cm of the facial vessels then that structure is not the marginal mandibular nerve (Stern 1992). A low cervical incision with a "nonvisualization" approach the marginal mandibular nerve is supported (Ichimura et al 1979, Smith et al 1993) for management of benign disease of the submandibular gland. In this study visualization of the nerve during the course of a neck dissection was not associated with a lower incidence of nerve dysfunction. The following technical considerations should minimize injury to the marginal mandibular nerve in the context of a neck dissection. Irrespective of the site of the skin incision skin flaps should be carefully elevated in a plane immediately deep to platysma and superficial to the investing layer of deep cervical fascia and marginal mandibular nerve. It is not the level of the skin incision that is important but the level of transection of investing layer of cervical fascia. Knowledge of the dynamic and variable location of the marginal mandibular nerve relative to the inferior border of the mandible is useful in this regard. The decision to visualize the nerve needs to be individualized (Stern 1992). If it is deemed necessary to divide the cervical fascia within 3-4 cm of the inferior border of the mandible it is our opinion that an attempt should be made to visualize the course of the nerve. The nerve is then preserved or sacrificed based on the oncologic objectives of the procedure. In our opinion electrical cautery should be used sparingly and attention should be directed to avoiding traction or pressure injury from retractors.

6.5 Resection Margin Status This historical cohort study of patients with squamous cell carcinoma of the oral cavity is relatively large. It is comparable in most respects, including the age and sex distribution, site distribution, etiologic factors, and incidence of second primaries, to most series of oral cancer. The incidence of involved margins (microscopic tumor at the inked resection margin) observed in this population based historical cohort was 14.6%. The reported incidence of involved margins following resection of head and neck cancers varies widely from 1 to 22% (Looser et al 1978, Chen et al 1987, Loree & Strong 1990, Jacobs et al 1993, Jones 1994, Spiro et al 1999, Sessions et al 2002, Sutton et al 2003, Kademani et al 2005). These studies with one exception (Jacobs et al 1993) report single institutional experience and often include extra- oral sites (Looser et al 1978, Chen et al 1987, Loree & Strong 1990, Jacobs et al 1993, Jones 1994). Margins are more often positive in oral cancer than other sites in the head and neck (Jones 1994, Woolgar & Triantafyllou 2005). Tumor at the inked resection margin increased recurrence rates in this study consistent with previous observations (Looser et al 1978, Loree & Strong 1990, Jacobs et al 1993, Spiro et al 1999, Sutton et al 2003). An independent and significant impact on overall survival was demonstrated. Involved margins were associated with an increased tumor load, advanced T and N status, on presentation. The association of involved margins with factors that can independently influence outcome in oral cancer such as stage of disease (Ravasz et al 1991, Sutton et al 2003, Ravasz et al 2005) and adverse histopathological features such as tumor thickness (Yuen et al 2000) and unfavourable patterns of invasion (Woolgar et al 1995, Sutton et al 2003, Kademani et al 2005) make the use of multivariate models necessary to assess the complex interactions of margin status and survival. Sutton et al (2003), in a detailed analysis of the impact of margin status on outcome, included clinical findings as well as histopathology data including nodal involvement and extracapsular spread, pattern of invasion, perineural invasion and vascular permeation in

their multivariate model. The relative risk of death for involved margins was 11.61(P=0.0013) and close margins 2.66(p=0.02). The only other factor with a significant impact on survival was positive lymph nodes (HR 2.15, p=0.063). Other studies have failed to show an independent effect of positive or inadequate margins on survival with multivariate analysis (Jones 1994, McMahon et al 2003, Kademani et al 2005). Theses studies are not strictly comparable to the above series as they included significant numbers of patients with extra-oral tumors (Jones 1994, McMahon et al 2003) previously treated patients (Jones 1994) and a different definition for the "positive margin" (McMahon et al 2003, Kademani et al 2005). Close margins (<2mm) had a similar impact on the incidence and pattern of local recurrence as involved margins. Survival was not adversely effected by the close margin. Many investigators consider both close, and tumor at the inked resection margins together in their analysis of the impact of margin status on outcome (Sadeghi et al 1986, Ravasz et al 1991, Woolgar et al 1995, van Es et al 1996, Slootweg 2002). When the impact of the close margin is reported separately the results suggest that the prognosis is better than observed in patients with tumor at the resection margin (Chen et al 1987, Loree & Strong 1990 Sutton et al 2003, Chandu et al 2005). A major problem in assessing and comparing the results of studies evaluating margins is the fact that some series include patients with close margins and others do not. This problem is further compounded by the lack of a standard definition for the close margin. The most widely accepted definition of a close margin is tumor within 5 mm of the inked resection margin (Looser et al 1978, Chen et al 1987, Loree & Strong 1990, Ravasz et al. 1991, Sutton et al 2003, Meier et al 2005). This is an arbitrary designation (Loree & Strong 1990, Batsakis JG 1999). Others use <2mm (Yuen et al 1998, Kademani et al 2005), <3mm (Chandu et al 2005), within one high power field (Spiro et al 1999) or a variable definition related to the pattern of invasion at the tumor host interface (Slootweg et al 2002). In-situ change and dysplasia have also been included ((Looser et al 1978, Sadeghi et al 1986, Chen et al 1987, Loree & Strong 1990, Ravasz et al 1991, Sutton et al 2003). This study did not address the impact of premalignant change at the margin. The balance of opinion would favor not including dysplasia in the definition (Meier et al 2005). The results of this study suggest that the outcome with tumor close to the resection margin should be reported separately from that observed with tumor at the inked resection margin as the prognosis differs. A standard definition for a close margin is necessary and this should be determined by a systematic evaluation of the effect of the width of resection on outcome. Considering the significant impact on outcome precise definitions for the inadequate surgical margin are important clinically. Patients with assumed inadequate margins are considered for adjunctive radiotherapy. It is accepted that adjunctive radiotherapy in the setting of an inadequate may improve local-regional control (Meier et al. 1984, Zelefsky et al 1992, Zelefsky et al 2004) and possibly impact survival (Sessions et al 1986, Sessions et al 2002). Adjunctive radiotherapy did not impact local and regional recurrence or survival in this study. It did however significantly increase morbidity.

The surgical margin is a prognostic determinant that is, at least in part, under the control of the surgeon as he or she determines the extent of resection in the individual patient. The clinical margin necessary to achieve an adequate histological margin has to take into consideration the fact that surgical specimens shrink, in the order of 40-50% (Batsakis 1999) when fixed in formalin. Mucosal margins tend to receive the most attention, however most recurrences involve the deep resection margin. The importance of a 3-dimensional resection has been emphasized in the work of Ravaszc and co-workers (Ravasz et al 1991). Recurrence was not identified in patients with positive mucosal margins. Positive deep and multiple positive margins were associated with recurrence rates of 38% and 70% respectively (Ravasz et al 1991). The positive association of margin status with various tumor related factors observed in this and other studies indicates that negative margins are harder to achieve with large tumor loads, proximity to mandible and a posterior location in the oral cavity. The positive association of margins with unfavorable histopathology such as tumor invading deeply as nests and cords as opposed to a pushing front has prompted several investigators (McMahon et al 2003, Sutton et al 2003) to consider positive margins as an indicator of more aggressive disease. Patients with poor histopathological parameters with clear margins have a better outcome than patients with bad histology and positive margins (Ravasz et al 1991, van Es et al 1996). The patient’s expectations for a functional outcome also influence the extent of resection. There is no single definition for an adequate clinical resection margin. As defined by Yuen and co-workers (Yuen et al 1998) the optimal resection margin should not compromise local control from an inadequate resection or cause unnecessary functional morbidity from too much resection. The use of intra-operative frozen section did not influence margin status in this study. Frozen section tends to emphasize mucosal and not the deep resection margin. In addition the common practice of assessing the patients resection margins (Yuen et al 2005), does not identify the patient with a close margin.

In summary the status of the surgical resection is an important predictor of outcome, both local recurrence and overall survival in oral cancer. The presence of tumor at the resection margin, and close to the resection margin should be considered separately with respect to prognostic significance. Intra-operative assessment of resection margins needs to emphasize involvement and proximity of tumor to the deep resection margin. There is a need for a standard definition for the close margin in oral cancer based on a systematic evaluation of the impact of the width of the margin on outcome.

6.6 Future Perspectives Improvements in cancer management will help increase survival. Screening has been shown to help reduce mortality in breast and colorectal cancer (Hakama et al 2008). Efficacy in mortality end-point reduction from screening has also been reported for hepatocellular carcinoma and oral cancer (Hakama et al 2008). Such screening efforts are even more important as some types of oropharyngeal carcinomas have been reported to be increasing (Hammarstedt et al 2007). Knowing that oral cancer detection at earlier stages is linked to improved survival adds more support to the importance of effective screening programs (Sargeran et al 2008).

The link to potential causative agents must also continue to receive attention. The finding that there is consistency in exposure intensity across diverse caner sites suggests a general phenomenon and may provide clues to the molecular basis of smoking-related cancer risk (Lubin et al 2008). Exposure to chemicals and alcohol also adds to the risk of oral cancer (Tarvainen et al 2008). The role of HPV as a risk factor for tonsillar cancer may also provide clues for etiology and guide prevention programs involving risk reduction and possible vaccination (Hammarstedt et al 2006).

Surgical treatment may also become more sophisticated. Resection margin analysis and monitoring may lead to better guidance of the surgeon by histopathologic results (Naison et al 2009). Immunostaining of markers such as claudin 7 could be used for prognostic purposes in patients with SCC of the tongue (Bello et al 2008). While such advances may be for sites other than the RMT or the maxillary alveolus and hard palate, they represent potential advances in the management of oral SCC, a disease with uncertain prognostic outcomes to the present day.

7 Summary and conclusions The following main conclusions can be drawn from the results obtained in the present studies:

1. The treatment outcomes of squamous cell carcinoma of the retromolar trigone using a population-based study resulted in an overall survival of 51.4% in this series. The optimal management of SCC of the RMT cannot be determined from this or any other single review. When considering relatively uncommon tumors such as SSC of the RMT there is a need for the head and neck oncology community to pool population based outcome data to establish accurate prognostic indicators.

2. Squamous cell carcinoma of the maxillary alveolar ridge and hard palate differs from other oral

cancers in several aspects. SCC tends to occur at these sites in a relatively older population with a female predilection. The treatment outcomes of SSC of the maxillary alveolus and palate using a population-based study showed that a relatively high number of patients were not amenable to treatment with curative intent. Treatment with surgery, with or without radiotherapy, appeared to improve disease control. A future prospective multi-centre study with a larger cohort of patients may help to answer questions regarding the prognosis of treatment related to the specific stages of this disease.

3. The relationship of the maxillary artery and the lateral pterygoid muscle in ethnic samples is

important in oral cancer care due to potential ethnic variations. These variations should be considered in planning of surgery in this anatomic region.

4. The anatomy and function of the marginal mandibular branch of the facial nerve pertains to the

surgical management of oral cancer as the function of this nerve may need to be sacrificed, causing significant facial disfigurement. Resections should be approached on a case-by-case basis as should the perceived need for the complete dissection and identification of the MMN. The goal should be to minimize disfigurement due to nerve palsy where possible.

5. The significance of the positive surgical margin in oral cancer management was assessed. The

status of the surgical resection is an important predictor of outcome, both local recurrence and overall survival in oral cancer. Intra-operative assessment of resection margins needs to emphasize involvement and proximity of tumor to the deep resection margin.

Acknowledgements This work was carried out under the auspices of Regea Institute for Regenerative Medicine, University of Tampere, Tampere, Finland. This work also involved my parent institutions in Canada, The Hospital for Sick Children, The Bloorview Kids Rehab, Cancer Care Manitoba, and the Department of Oral & Maxillofacial Surgery both at the University of Manitoba and the University of Toronto.

I wish to express my deepest gratitude and thanks to my mentor and supervisor and my very dear friend of many years Professor George K Sándor, M.D., D.D.S., Ph.D., F.R.C.D.(C), F.R.C.S.(C), F.A.C.S., Professor of Tissue Engineering, Regea Institute for Regenerative Medicine, University of Tampere, Tampere, Finland, Professor and Head of Oral and Maxillofacial Surgery, University of Toronto, Coordinator of Pediatric Oral and Maxillofacial Surgery, The Hospital for Sick Children and Bloorview Kids Rehab, Toronto, Canada and Dosent in Oral and Maxillofacial Surgery, University of Oulu, Oulu, Finland for providing me with his expert guidance, meaningful criticisms, support and special friendship during these many long years. His never ending enthusiasm and encouragement has been extremely valuable for the commencement and completion of this thesis. Professor Sándor made this work possible!

I wish to express my thanks and gratitude to my co-supervisor, Professor Riitta Suuronen, M.D., D.D.S., Ph.D., B.V.M., F.E.B.O.M.F.S., Director of Regea Institute for Regenerative Medicine, University of Tampere, Tampere, Finland for her support and help with making this thesis possible.

I would like to thank my close friend and partner, Associate Professor Richard Nason M.D., F.R.C.S., Head Department of Surgery, University of Manitoba, for giving me the continuous support through out my training and research activity.

I express my warmest thanks for the careful revisions and valuable criticism of the manuscript to the official referees, Dosent Kalle Aitesalo, M.D., Ph.D., Otolaryngology, Head and Neck Surgery and Oral and Maxillofacial Surgery, University of Turku, Turku, Finland and Dosent Jyrki Törnwall M.D., D.D.S., Ph.D., Head of Oral and Maxillofacial Surgery, Helsinki University Hospital, Helsinki, Finland. Your thoughtful suggestions have gone a long way to improve the quality of this thesis.

I wish to thank the co-authors of the five publications which form the nucleus of this thesis, Richard W. Nason M.D., F.R.C.S.C., Ahmed A Abdoh Ph.D., Ali Hussain B.Sc., D.D.S., F.R.C.D., Algernoun Karim Ph.D., Mark G Torchia Ph.D. and James Thliversis Ph.D. It has been a pleasure to collaborate with you all!

I wish to thank my present and former colleagues and other staff in the Departments of Oral and Maxillofacial Surgery both at the Universities of Manitoba and Toronto for their congenial support and friendship over these many years. Thank you to Associate Professor John B. Curran B.D.S., F.F.D.R.C.S.(IREL), F.R.C.D.(C), Head of the Department of Oral and Maxillofacial Surgery at the University of Manitoba, Assistant Professor Leland R. McFadden D.D.S., M.Sc., F,R.C.D.(C) and Mark A. Cohen B.D.S., F.R.C.D.(C) for their instruction during my residency. I would like to express very special gratitude to my very dear professional friend Assistant Professor Robert P. Carmichael, D.M.D., M.Sc., F.R.C.D.(C), Chief of Dentistry, Bloorview Kids Rehab, Toronto, Canada, who was able to help me as a friend and confidant over these years of study. Thank you! I would also thank Professor Cameron M.L. Clokie D.D.S., Ph.D., F.R.C.D.(C) for his encouragement.

I would like to thank all those responsible for my training as a specialist, to those at the Department of Oral and Maxillofacial Surgery, University of Manitoba and especially to those in the Division of Surgical Oncology, University of Manitoba, Department of Oral and Maxillofacial Surgery University of Toronto, Canada.

I would like to express my appreciations to King Abdullah International Medical Research Center for the support that they provided me with to finalize this work. Also special gratitude is owing to King Abdulaziz Medical City in the Kingdom of Saudi Arabia, who sponsored me during my training and gave me the chance to be here in Finland now.

I wish to express my sincerest apprec iation to the People of Finland for allowing me the opportunity to complete this work at one of their very fine universities, University of Tampere. This occasion gives me the opportunity as a Saudi to publicly thank your country for its special friendship and all of its support shown to my country.

I owe a great deal of loving thanks to my mother and father for all their support and encouragement during these many years. They were always there to be the best possible parents and examples of boundless honest hard work to me. They constantly encouraged me to study hard with vigour and enthusiasm. I owe them my deepest gratitude. I also would like to express my deepest thanks to the Soul of my father-in law who passed away on February 16th of 2009, I have learned from him that hard work and patience are the road to success.

Finally, I wish to express my loving thanks to my wife, Munira Alhugail, who stood next to me during all the hard years of training, she tolerated my absences, and supported me to reach my goals, Thank you. To my two very special children, Haya and Albandry, Thank you.

Tampere, May 2009 Abdulaziz Binahmed

8 References Ackerman LV. Verrucous carcinoma of the oral cavity. Surgery 1948; 23: 670-678. Adachi B: Das Arteriensysteem der Japaner, Kyoto, Maruzen, vol 1, 1928. Adair TH, Gay WJ and Montani JP (1990). Growth regulation of the vascular system: Evidence for metabolic hypothesis. Am J Physiol 259: R393-R404. Allen WE, Kierl EL, and Rothmann SLG (1974). The maxillary artery in craniofacial pathology. Am J Roentgenol 121:124-138. Amdur RJ, Parsons JT, Mendenhall WM, Million RR, Stringer SP and Cassisi NJ (1989). Postoperative irradiation for squamous cell carcinoma of the head and neck: an analysis of treatment results and complications.Int J Radiat Oncol Biol Phys 16: 25-36. Antoniadoes K, Lazardidis N, Vahtsevanos K, Hadjipetrou L, Antoniadoes V and Karakasis D. (2003). Treatment of squamous cell carcinoma of the anterior faucial pillar–retromolar trigone. Oral Oncol 39: 680-686. Apinhsmit W, Methathrathip D, Ploytubtim S, Chompoopong , Ariyawatkul T and Lertsirithong A (2004). Anatomical study of the maxillary artery at the pterygomaxillary fissure in a Thai population; its relationship to maxillary osteotomy. J Med Assoc Thailand 87: 1212-1217. Ayad T, Gelinas M, Guertin L, Larochelle D, Vecchio P and Tabet JC (2005). Retromolar trigone carcinoma treated by primary radiation therapy: an alternative to the primary surgical approach. Arch Otolaryngol Head Neck Surg;131:576-582. Baker DC and Conley J (1979). Avoiding facial nerve injuries in rhytidectomy. Plast Reconstr Surg: 64: 781–795. Baker SR (1993). Malignant neoplasms of the oral cavity. In Cummings CW (ed). Otolaryngology: Head and Neck Surgery. St. Louis: Mosby 1248-1305. Bello IO, Vilen ST, Niinimaa A, Kantola S, Soni Y and Salo T (2008). Expression of claudins 1, 4, 5, and 7 and occluding, and relationship with prognosis in squamous cell carcinoma of the tongue. Human Pathol 39: 1212-1220. Barbosa J (1959). Cancer of the retromolar trigone. Arch Otolaryngol 69: 19-30.

Batsakis JG (1999). Surgical excision margins: a pathologist's perspective. Adv Anat Pathol. 6:140 148. Beaumont DG and Hains JD. Change in surgical margins in vivo following resection and after fixation. Aust J OtoLaryngol 1992; 1: 51-2 Beitler JJ, Smith RV, Silver CE, Quish A, Deore SM, Mullokandov E, Fontenla DP, Wadler S, Hayes MK and Vikram B (1998).Close or positive margins after surgical resection for the head and neck cancer patient: the addition of brachytherapy improves local control. Int J Radiat Oncol Biol Phys 40: 313-317. Bhansali S (1961). Malignant tumors of the buccal cavity: a clinical analysis of 970 cases. Clin Radiol 12: 299-308. Block G, Patterson B and Subar A (1992). Fruit, vegetables, and cancer prevention: a review of the epidemiological evidence. Nutr Cancer 18: 1-29. Boffetta P, Mashberg A, Winkelmann RV and Garfinkel L (1992). Carcinogenic effect of tobacco smoking and alcohol drinking on anatomic sites of the oral cavity and oropharynx. Int J Cancer 52: 530-533. Brown RL, Shu JM, Scarborough JE, Wilkins SA and Smith RR (1965). Snuff dipper’s intraoral cancer: clinical characteristics and response to therapy. Cancer 18: 2-13. Byers R, Anderson B and Schwars E (1984). Treatment of squamous cell carcinoma of the retromolar trigone. Am J Clin Oncol 7: 647-652. Byers RM, Bland KI, Borlase B and Luna M (1978). The prognostic and therapeutic value of frozen section determinations in the surgical treatment of squamous carcinoma of the head and neck Am J Surg 136: 525-528. Cady B and Catlin D (1969). Epidermoid carcinoma of the gum. A 20-year survey. Cancer 23: 551 569. Canadian cancer statistics (2004). http://www.cancer.ca/Canada-wide/About%20cancer/Types Carmichael FA and McGowan DA (1992). Incidence of nerve damage following third molar removal: a West of Scotland Oral Surgery Research Group study. Br J Oral Maxillofac Surg 30: 78-82. Chandu A, Adams G and Smith ACH (2005). Factors affecting survival in patients with oral cancer: an Australian perspective. Int J Oral & Maxillofac Surg 34: 514-520. Chen TY, Emrich LJ and Driscoll DL (1987). The clinical significance of pathological findings in surgically resected margins of the primary tumor in head and neck carcinoma. Int. J. Radiation Oncology Biol Phys 1987 3: 833-837. Chung K, Rahman S, Lim M and Constable W (1979). Squamous cell carcinoma of the hard palate. Int J Radiation Oncology Biol. Phys 5: 191-196. Chung C, Johns M, Cantrell R and Constable W (1980). Radiotherapy in the management of primary malignancy of the hard palate. Laryngoscope 90: 576-584. Coleman RD and Smith RA (1982). The anatomy of mandibular anesthesia: Review and analysis. Oral Surg Oral Med Oral Path Oral Radiol Endod 54: 148-152.

Cox DR (1972). Regression models and life tables. J R Stat Soc 34: 187-189. Dingman RO & and Grabb WC (1962). Surgical anatomy of the mandibular ramus of the facial nerve based on the dissection of 100 facial halves. Plast Reconstr Surg: 29: 266–272. Evan J and Shah J (1981). Epidermoid carcinoma of the palate. Am J of Surg. 142: 4551-4555. Fisch U (1978). Infratemporal fossa approach to tumors of the temporal bone and base of the skull. J Laryngol Otol 92: 949-967. Fisch U (1982). Infratemporal fossa approach for glomus tumors of the temporal bone. Ann Otol Rhinol Laryngol 91: 474-479. Fisch U and Pillsbury HC (1997). Infratemporal fossa approach to lesions in the temporal bone and base of the skull. Arch Otolaryngol 105: 99-107. Franceschi S, Barra S, La Vecchia C, Bidoli E, Negri E and Talamini R (1992). Risk factors for cancer of the tongue and the mouth. A case-control study from northern Italy. Cancer 70: 2227-2233. Franceschi D, Gupta R, Spiro RH and Shah IP (1993). Improved survival in the treatment of squamous cell carcinoma of the tongue. Am J Surg 166: 360-365. Frangiskos F, Stavrou E, Tsitsogianis H, Vardas E and Antonopoulou I (2003). Incidence of penetration of a blood vessel during inferior alveolar nerve block. Br J Oral and Maxillofacial Surg 41: 188 189. Fujimura k, Segami N and Kobayahi S (2006). Antomical study of the complications of intraoral vertico-sagittal ramus osteotomy. J Oral Maxillofac Surg 64: 384-389. Gates GA (1988). The lateral facial approach to the nasopharynx and infratemporal fossa. Otolaryngol Head Neck Surg 99: 321-325. Genden E, Ferlito A, Shaha A and Rinaldo A (2003). Management of cancer of the retromolar trigone, Oral Oncol 39: 633-637. Gomez D, Faucher A, Picot V, Siberchoicot F, Renaud-Salis JL, Bussieres E and Pinsolle J (2000). Outcome of squamous cell carcinoma of the gingiva: a follow-up study of 83 cases. J Craniomaxillofac Surg 28: 331-335. Gupta PC, Mehta FS, Daftary DK, Pindborg JJ, Bhonsle RB, Jalnawalla PN, Sinor PN, Pitkar VK, Murti PR, Irani RR, Shah HT, Kadam PM, Iyer KS, Iyer HM, Hegde AK, Chandrashekar GK, Shiroff BC, Sahiar BE and Mehta MN (1980). Incidence rates of oral cancer and natural history of oral precancerous lesions in a 10-year follow-up study of Indian villagers. Community Dent Oral Epidemiol 8: 283-333. Hakama M,Coleman MP,Alexe DM, and Auvinen A (2008). Cancer screening: evidence and practice Europe in 2008. Eur J Cancer 44: 1404-1413. Hald J and Andreassen UK (1994). Submandibular gland excision: short- and long-term complications. J Otorhinolaryngol Relat Spec 56: 87–91. Hammerstedt L, Lindquist D, Dahlstrand H, Romanitan M, Dahlgren LO, Joneberg J, Creson N, Lindholm J, Ye W, Dalianis T and Munck-Wikland E (2006). Human papillomavirus as a risk factor for the increase in

incidence of tonsillar cancer. Int J Cancer 119: 2620-2623. Hammerstedt L, DahlstrandH, Lindquist D, Onelöv L, Ryott M, Luo J, Dalianis T, Ye W and Munck- Wikland E (2007). The incidence of tonsillar cancer in Sweden is increasing. Acta Otolaryngol 2007: 988- 992. Hao SP, Tsang NM, Chang KP, Chen CK and Huang SS (2006). Treatment of squamous cell carcinoma of the retromolar trigone. Laryngoscope. 116: 916-920. Haskell R (1986). Medico-legal consequences of extracting lower third molar teeth. Medical Protection Society annual reports and accounts 94: 51-52. Hicks WL Jr, Loree TR, Garcia RI, Maamoun S, Marshall D, Omer JB, Bakamjian VY and Shedd DP (1997). Squamous cell carcinoma of the floor of the mouth: a 20 year review. Head and Neck 19: 400-405. Holliday MJ (1986). Lateral transtemporal-sphenoid approach to the skull base. Ear Nose Throat 65: 153-162. Huang C, Chao K, Tsai J, Simpson JR, Haughey B, Spector GJ and Sessions DG (2001). Cancer of the retromolar trigone: long-term radiation therapy outcome. Head and Neck Surg 23: 758-763. Ichimura K, Nibu K-I and Tanaka T. (1997) Nerve paralysis after surgery in the submandibular triangle: review of University of Tokyo Hospital experience. Head & Neck 19:48-53. International Agency for Research on Cancer. Tobacco smoking (1986). IARC Monographs on the evaluation of carcinogenic risks to humans. IARC 38: 400. International Union Against Cancer. TNM Classification of Malignant Tumors 6th ed (2002). New York: Wiley-Liss 22-26. Jacobs JR, Ahmad K, Casiano R, Schuller DE, Scott C, Laramore GE and Al-Sarraf M (1993). Implications of positive surgical margins. Laryngoscope 103: 64-68. Jemal A, Murray T, Ward E, Samuels A, Tiwari RC, Ghafoor A, Feuer EJ and Thun MJ (2005) Cancer statistics, 2005. CA Cancer J Clin 55: 10-30. Johnson NW (1997). Oral cancer: a world wide problem. FDI World 6: 19-21. Johnson RE, Sigman JD, Funk GF, Robinson RA and Hoffman HT (1997). Quantification of surgical margin shrinkage in the oral cavity. Head Neck. 19: 281-286. Johnston WD and Ballantyne AJ (1977). Prognostic effect of tobacco and alcohol use in patients with oral tongue cancer. Am J Surg 134: 444-447. Jones AS (1994). Prognosis in mouth cancer: tumour factors. Oral Oncol Eur J Cancer 30B: 8115. Kademani D, Bell, RB, Bagberi S, Homgren E, Dierks E, Potter B and Homer L (2005). Prognostic factors in intraoral squamous cell carcinoma: The influence of histologic grade. J Oral Maxiollofac Surg 63: 1599-1605. Kaplan ER and Meier P (1958). Non-parametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481.

Kari S, Alho OP, Jokinen K, Hyrynkangas K, Läära E (1997). Carcinoma of the tongue in Northern Finland: trends in overall incidence and patient and tumour characteristics. J Oral Pathology and Medicine 26: 480-483. Kato I and Nomura AM (1994). Alcohol in the aetiology of upper aerodigestive tract cancer. Eur J Cancer B Oral Oncol 30B: 75-81. Kim MS, Shin KH, Baek JH, Cherrick HM and Park NH (1993). HPV-16, tobacco-specific N nitrosamine, and N-methyl-N'-nitro-N-nitrosoguanidine in oral carcinogenesis. Cancer Res 53: 4811-4816. Kowalski L, Hashimoto I and Magrin J (1993). End results of 114 extended “commando” operation for the retromolar trigone carcinoma. Am J Surg 166: 374-379. Krizan Z (1960). Beitrage zur deskriptiven und topographischen Anatomie der A. maxillaries. Acta Anat. 41:319-333. Krolls S and Hoffman S (1976). Squamous cell carcinoma of the oral soft tissues: a statistical analysis of 14,253 cases by age, sex, and race of patients. J Am Dent Assoc 92: 571-574. Krubhaar EB (1937). Clio Med 19, Pathology. New York: PB Hoeber, Inc 171. La Vecchia G and Tavani A (1998). Fruit and vegetables and human cancer. Eur J Cancer Prev 7: 3-8. Laskar GW, Opdyke DL and Miller H (1951). The position of the internal maxillary artery and its questionable relation to the cephalic index. Anat Rec 109: 119-126. Lauber H (1901). Uber einige Varietaten im Verlauf der Arteria maxillaries interna. Anat Anz 19: 444-448. Lehman RH, Cox ID, Belson TP, Yale RS, Byhardt RW, Toohill RI and Malin TC (1982). Recurrence patterns by treatment modality of carcinomas of the floor of the mouth and tongue. Am J Otolaryngol 3: 174-181. Lewin F, Norell SE, Johansson H, Gustavsson P, Wennerberg J, Biorklund A and Rutqvist LE (1998). Smoking tobacco, oral snuff, and alcohol in the etiology of squamous cell carcinoma of the head and neck: a population-based case-referent study in Sweden. Cancer. 82: 1367-1375. Llewellyn CD, Johnson NW and Warnakulasuriya KA (2001). Risk factors for squamous cell carcinoma of the oral cavity in young people--a comprehensive literature review. Oral Oncol 37: 401-418. Lo K, Fletcher G and Byers R (1987). Results of irradiation in the squamous cell carcinoma of the anterior faucial pillar–retromolar trigone. Int J Radiat Oncol Biol Phys 13: 969-974. Looser K, Shah J and Strong E (1978). The significance of positive margins in surgically resected epidermoid carcinoma. Head and Neck Surg 1: 107-111. Loree TR and Strong EW (1990). Significance of positive margins in oral cavity squamous carcinoma. .Am J Surg. 160: 410-414. Lubin JH, Virtamo J, Weinstein SJ and Albanes D (2008). Cigarette smoking and cancer: intensity patterns in the alpha-tocopherol, beta-carotene cancer prevention study in Finnish men. Am J Epidemiol 167: 970- 975.

Lurje A (1946). On topographical anatomy of the inner maxillary artery. Acta Anat 2: 219-231. Macho GA (1986). Cephalometric and craniometric age changes in adult human. Ann Human Biol 13: 49-61. Makitie AA, Lundberg M, Salmela K, Kyllonen L and Pukkala E (2008). Head and neck cancer in renal transplant patients in Finland. Acta Otolaryngol 128: 1255-1258. Mao EJ (1995). Prevalence of human papillomavirus 16 and nucleolar organizer region counts in oral exfoliated cells from normal and malignant epithelia. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 80: 320-329 McCarty PJ and Million RR (1994). History of diagnosis and treatment of cancer in the head and neck. Management of Head and Neck Cancer: A Multidisciplinary approach, 2nd ed. RR Million and NJ Cassis (eds) PP.1-29. JB Lippincott: Philadelphia. McCoy JM and Waldron CA (1981).Verrucous carcinoma of the oral cavity. A review of forty-nine cases. Oral Surg Oral Med Oral Pathol 52: 623-629 McGregor I and McGregor F (1986). Cancer of the Face and Mouth. McCoy JM & Waldron CA (eds). Pathology and Management for Surgeons. Edinburgh: Churchill Livingstone 499-510. McLaughlin JK, Gridley G, Block G, Winn DM, Preston-Martin S, Schoenberg JB, Greenberg RS, Stemhagen A, Austin DF and Ershow AG (1988). Dietary factors in oral and pharyngeal cancer.J Natl Cancer Inst 80: 1237-1243. McMahon J, O'Brien CJ, Pathak I, Hamill R, McNeil E, Hammersley N, Gardiner S and Junor E (2003). Influence of condition of surgical margins on local recurrence and disease-specific survival in oral and oropharyngeal cancer. Br J Oral Maxillofac Surg 41: 224-231 Meier JD, Oliver DA and Varvares MA (1984). Failure at the primary site following multimodality treatment in advanced head and neck cancer.Head Neck Surg. 6: 720-723. Meier JD, Oliver DA and Varvares MA (2005). Head Neck 27: 952-958. Surgical margin determination in head and neck oncology: Current clinical practice. The results of an International American Head and Neck Society Member Survey. Mendenhall WM, Morris CG, Amdur RJ, Werning JW and Villaret DB (2005). Retromolar trigone squamous cell carcinoma treated with radiotherapy alone or combined with surgery. Cancer 103: 2320-2325. Mickey B, Close L, Schaefer S and Samson D (1988). A combined fronto-temporal and lateral infratemporal fossa approach to the skull base, J Neurosurg 68: 678-683. Moffat DA and Ramsden RT (1977). The deformity produced by a palsy of the marginal mandibular branch of the facial nerve. J Laryngol Otol 91: 401–406. Moore SR, Johnson NW, Pierce AM and Wilson DF (2000a). The epidemiology of mouth cancer: a review of global incidence Oral Dis 6: 65-74. Moore SR, Johnson NW, Pierce AM and Wilson DF (2000b). The epidemiology of tongue cancer: a review of global incidence Oral Dis 6: 75-84.

Nason RW, Binahmed A Pathak KA, Abdoh AA, and Sándor GKB (2009). What is the adequate margin of resection in oral cancer? Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology. Epub January 23. Nelson DW and Gingrass RP (1979). Anatomy of the mandibular branches of the facial nerve. Plast Reconstr Surg 64: 479-482. New G and Hallberg O (1941). The end results of the treatment of malignant tumors of the palate. Surg Gynecol Obstet 73: 520-524 Oliver AJ, Helfrick JF and Gard D (1996). Primary oral squamous cell carcinoma: a review of 92 cases. J Oral Maxillofac Surg. 54: 949-954. Ong J (1970). Cancer of the oral and pharyngeal cavities. J R Coll Surg Edinb 15: 250-267. Orbay H, Unlu RE, Sensoz O and Comert A (2006). Vascular anatomy of the lateral pterygoid muscle. Ann Plast Surg 56: 584-585. Ord RA and Aisner S (1997). Accuracy of frozen sections in assessing margins in oral cancer resection. J Oral Maxillofac Surg 55: 663-669. Peng J, Raverdy N, Ganry O, de La Roche-Saint-Andre G, Dubreuil A and Lorriaux A (2000). Descriptive epidemiology of upper aerodigestive tract cancers in the department of Somme. Bull Cancer 87: 201-206 Petruzzelli G and Mayers E (1994). Malignant neoplasms of the hard palate and upper alveolar ridge. Oncology 8: 43-48. Pisani P, Parkin DM and Ferlay J (1993). Estimates of the worldwide mortality from eighteen major cancers in 1985. Implications for prevention and projections of future burden. Int J Cancer 55: 891-903. Potter JD and Steinmetz K (1996). Vegetables, fruits and phytoestrogens as prevwentive agents. IARC Sci Publ 139: 69-90. Pretterklieber ML, Skopakoff C and Mayr R (1991). The human maxillary artery reinvestigated: topographical relations in the infratemporal fossa. Acta Anat 142: 281-287. Ramulu C and Reddy C (1972). Carcinoma of the hard palate and its relation to reverse smoking. Int Surg 57: 636-640. Ranasinghe AW, Johnson NW and Scragg MA (1978). Iron deficiency depresses cell proliferation in hamster cheek pouch epithelium. Cell Tissue Kinet 20: 403-412. Ravasz LA, Slootweg PJ, Hordijk GJ, Smit F and van der Tweel I (1991). The status of the resection margin as a prognostic factor in the treatment of head and neck carcinoma. J. Cranio Max Fac Surg 19: 314-318. Ravasz LA, Slootweg PJ, Hordijk GJ, Smit F and van der Tweel I (2005). A histopathological appraisal of surgical margins in oral and oropharyngeal cancer resection specimens. Oral Oncology 41: 1034-1043. Reddy C. (1974). Carcinoma of hard palate in India in relation to reverse smoking of chuttas. J Natl Cancer Inst 53: 615-619.

Rennie JS and MacDonald DG (1984). Cell kinetics of hamster ventral tongue epithelium in iron deficiency. Arch Oral Biol 29: 195-199. Sadeghi A, Kuisk H, Tran LM, Mackintosh R, Mclaren JR and Parker RG (1986). The role of radiation therapy in squamous cell carcinoma of the upper aerodigestive tract with positive surgical margins. Am J Clin Oncol CCT 9: 500-503. Sargeran K, Murtomaa H, Safavi SM, Vehkalahti MM and Teronen O (2008). Survival after diagnosis of cancer of the oral cavity. Br J Oral Maxillofac Surg 46: 187-191. Sasco AJ, Secretan MB and Straif K (2004). Tobacco smoking and cancer: a brief review of recent epidemiological evidence. Lung Cancer 45: Suppl 2 S 3-9. Savary V, Robert R, Rogez JM, Armstrong O and Leborgne J (1997). The mandibular marginal ramus of the facial nerve: an anatomic and clinical study. Surg Radiol Anat 19: 69-72 Schildt EB, Eriksson M, Hardell L and Magnuson A (1998). Oral infections and dental factors in relation to oral cancer: a Swedish case--control study. Eur J Cancer Prev 7: 201-206. Schlecht NF, Pintos J, Kowalski LP and Franco EL (2001). Effect of type of alcoholic beverage on the risks of upper aerodigestive tract cancers in Brazil. Cancer Causes Control 12: 579-587. Sessions DG, Spector GJ, Lenox J, Haughey B, Chao C and Marks J (1986). The role of radiation therapy in squamous cell carcinoma of the upper aerodigestive tract with positive surgical margins. Am J Clin Oncol(CCT) 9: 500-503. Sessions DG, Spector GJ, Lenox J, Haughey B, Chao C and Marks J (2002). Analysis of treatment results for oral tongue cancer. Laryngoscope 112: 616-625. Shah J. Global epidemiology,Oral cancer 2003 PP3-32 Shaha AR, Spiro RH, Shah IP and Strong EW (1984). Squamous carcinoma of the floor of the mouth. Am J Surg 148: 455-459. Shibuya H, Horiuchi J, Suzuki S, Amagasa M and Mashima K (1984). Oral carcinoma of the upper jaw. Results of radiation treatment. Acta Radiol Oncol. 23: 331-335. Shin KH, Tannyhill RJ, Liu X and Park NH (1996). Oncogenic transformation of HPV- immortalized human oral keratinocytes is associated with the genetic instability of cells. Oncogene 12: 1089 -1096. Silverman S Jr and Griffith M (1972). Smoking characteristics of patients with oral carcinoma and the risk for second oral primary carcinoma. J Am Dent Assoc 85: 637-640. Sjovold T (1988). Geschlechtsdiagnose am Sklett; in Knussmann R (ed): Anthropologie. Handbuch der vergleichenden Biologie des Menschen. Stuttgart, Fischer, vol 1, pp 444-479. Skandalakis JE, Gray SW and Rowe JS (1979). Surgical anatomy of the submandibular triangle. The American Surgeon 590-596. Skopakoff C (1968). Uber die variabilitat im verlauf der arteria maxillaries. Anat Anz 123: 531 546.

Slootweg PJ, Hordijk GJ, Schade, van Es RJJ and Koole R (2002). Treatment failure and margin status in head and neck cancer. A critical view on the potential value of molecular pathology. Oral Oncology 38: 500-503. Smith WP, Peters WJN and Markus AF (1993). Submandibular gland surgery: an audit of clinical findings, pathology and postoperative morbidity. Ann Royal Coll Surg Engl 75: 164–167. Soo K, Spiro R, King W, Harvey W, and Strong E (1988). Squamous cell carcinoma. Am J of Surg. 156: 281-285. Spiro RH, Guillamondegui O Jr., Paulino AF and Huvos AG (1999). Pattern of invasion and margin assessment in patients with oral tongue cancer. Head Neck 21: 408-413. Spruance SL, Reid JE, Grace M and Samore M (2004). Hazard ratio in clinical trials. Antimicrob Agents Chemother 48: 2787–2792. Sterling TD, Rosenbaum WL and Weinkam JJ (1992). Analysis of the relationship between smokeless tobacco and cancer based on data from the National Mortality Followback Survey. J Clin Epidemiol 45: 223-231. Stern S (1992). Anatomic correlates of head and neck surgery. Head & Neck 14: 328-331 Sutton DN, Brown JS, Rogers SN, Vaughan ED and Woolgar JA (2003). The prognostic implications of the surgical margin in oral squamous cell carcinoma. Int J Oral Maxillofac Surg 32: 30-34. Syrjänen KJ &\ and Surjänen SM (1981). Histological evidence for the presence of condylomatous epithelial lesions in association with laryngeal squamous cell carcinoma. J Otorhinolaryngol Relat Spec 43: 181-194. Tarvainen L, Kyyrönen P, Kauppinen T and Pukkala E (2008). Cancer of the mouth and pharynx, occupation and exposure to chemical agents in Finland in 1971 to 1995. Int J Cancer 123: 653-659. Thomas S and Wilson A (1993). A quantitative evaluation of the aetiological role of betel quid in oral carcinogenesis. Eur J Cancer B Oral Oncol 29B: 265-271. Thompson A (1891). Report of the committee of collective investigation of the Anatomical Society of Great Britain and Ireland for the year 1889-1890. J Anat 25: 89-101. Törnwall J, Snäll J and Mesimäki K (2008). A rare case of spinal cord metastases from oral SCC. Br J Oral Maxillofac Surg 46: 594-595. Tulley P, Webb A, Chana JS, Tan T, Hudson D, Grobbelaar AO and Harrison DH (2000). Paralysis of the marginal Mandibular branch of the facial nerve: treatment options. Br J Plast Surg 53: 378-385. U.K. Oral Cancer incidence statistics (2009). http://info.cancerresearchuk.org/cancerstats/types/oral/incidence/ Van Es RJJ, van Nieuw Amerongen N, Slootweg PJ and Egyedi P (1996). Resection margin as a predictor of recurrence at the primary site for T1 and T2 oral cancers. Arch Otolaryngol Head Neck Surg 122: 521-525. Ventä I and Lindqvist C (1998). Malpractice claims for permanent nerve injuries related to third molar removals Acta Odontol Scand 56: 193-196.

Vertcauteeren M (1990). Age effects and secular trend in a cross-sectional sample: Applications to forehead dimensions in Belgian adults. Hum Biol 62: 681-688. Vincent RG and Marchetta F. The relationship of the use of tobacco and alcohol patients to cancer of the oral cavity, Pharynx or larynx . Am J Surg. 1963 Sep;106:501-5. Vrionis FD, Cano WG and Heilman CB (1996). Microsurgical anatomy of the infratemporal fossa as viewed laterally and superiorly. Neurosurgery 39 (4): 777-785. Winn DM (1995). Diet and nutrition in the etiology of oral cancer.Am J Clin Nutr 61: 437S-445S. Woods KV, Shillitoe EJ, Spitz MR, Schantz SP and Adler-Storthz K (1993). Analysis of human papillomavirus DNA in oral squamous cell carcinomas.J Oral Pathol Med 22: 101-108. Woolgar JA, Brown JS, Scott J, West CR, Vaughan ED and Rogers S (1995). Survival, metastasis and recurrence of oral cancer in relation to pathological features. Ann R Coll Surg Engl 77: 325-331 Woolgar JA, Rogers S, West CR, Errington RD, Brown JS and Vaughan ED (1999). Survival and patterns of recurrence in 200 oral cancer patients treated by radical surgery and dissection. Oral Oncol 35: 257-265. Woolgar JA and Triantafyllou A (2005). A histopathological appraisal of surgical margins in oral and oropharyngeal cancer resection specimens. Oral Oncol 41: 1034-1043. Wunsch-Filho V and de Camargo EA (2001). The burden of mouth cancer in Latin America and the Caribbean: epidemiologic issues. Semin Oncol. 28: 158-168. Wynder EL, Mushinski MH, Stellman SD and Choay P (1977). Tobacco and alcohol consumption in relation to the development of multiple primary cancers. Cancer 40(4 Suppl): 1872-1878. Yeudall WA (1992). Human papilloma viruses and oral neoplasia. Eur J Cancer B Oral Oncol 28B: 61-66. Yokoo S, Umeda M, Komatsubara H, Shibuya Y and Komri T (2002). Evaluation of the T- classification of the upper gingival and hard palate carcinomas- a proposition for new criterion of T4. Oral Oncol 38: 378-382. Yorozu A, Sykes A and Slevin N (2001). Carcinoma of the hard palate treated with radiotherapy: a retrospective review of 31 cases. Oral Oncol 37: 493-497. Yuen PW, Lam KY and Wei WI (2000). A comparison of the prognostic significance of tumor diameter, length, width, thickness, area, volume, and clinicopathological features of oral tongue carcinoma. Am J Surg 180: 139-143. Yuen PW, Lam KY, Chan ACL, Wei WI and Lam LK (1998). Clinicopathological analysis of local spread of carcinoma of the tongue. Am JSurg 175: 242-244. Yuen PW, Lam KY, Chan ACL, Wei WI and Lam LK (2005). Surgical margin determination in head and neck oncology: Current clinical practice. The results of an international American head and neck society member survey. Head Neck 7: 952-958. Zelefsky MJ, Harrison LB, Fass DE, Armstrong JG, Shah JP and Strong EW (1992). Postoperative radiation therapy for squamous cell carcinomas of the oral cavity and oropharynx: Impact of therapy on patients with positive surgical margins. Int J Radiation Oncology Biol Phys 25: 17-21.

Zelefsky MJ, Harrison LB, Fass DE, Armstrong JG, Shah JP and Strong EW (2004). Postoperative irradiation for squamous cell carcinoma of the oral cavity: 35-year experience. Head Neck 26: 984-994. Zelefsky MJ, Harrison LB, Fass DE, Armstrong JG, Shah JP and Strong EW (1993). Postoperative radiation therapy for squamous cell carcinomas of the oral cavity and oropharynx: impact of therapy on patients with positive surgical margins. Int J Radiat Oncol Biol Phys 25: 17-21. Ziarah HA and Atkinson ME (1981). The surgical anatomy of the mandibular distribution of the facial nerve. Brit J Oral Surgery 19:159-170.

9 Role in publications List of papers:

I. Binahmed A, Nason RW, Abdoh AA, Sándor GKB (2007). Population-based study of treatment outcomes in squamous cell carcinoma of the retromolar trigone. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology. 104: 662-665. Epub 2007 Aug 29. Role: Collected and analyzed data, statistical analysis of results with the help of Dr. Abdoh, wrote paper.

II. Binahmed A, Nason RW, Hussain A, Abdoh AA, Sándor GKB (2008). Treatment outcomes in

squamous cell carcinoma of the maxillary alveolus and palate: A population-based study. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology. 105: 750-754. Epub 2008 Feb 23. Role: Collected and analyzed data, statistical analysis of results with the help of Dr. Abdoh, wrote paper.

III. Hussain A, Binahmed A, Karim A, Sándor GKB (2008). Relationship of the maxillary artery and lateral pterygoid muscle in a caucasian sample. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontology. 105: 31-35. Epub 2007 Aug 29. Role: Performed dissections, collected data, statistical analysis and wrote paper.

IV. Nason RW, Binahmed A, Torchia MG, Thliversis J (2007). Clinical observations of the

anatomy and function of the marginal mandibular nerve. Int J Oral Maxillofac Surg. 36: 712-715. Epub 2007 March 27. Role: Performed dissections, collected data, statistical analysis and wrote paper

V. Binahmed A, Nason RW, Abdoh AA (2007). The clinical significance of the positive surgical

margin in oral cancer. Oral Oncol. 43: 780-784. Role: Collected and analyzed data, statistical analysis of results with the help of Dr. Abdoh, wrote paper.

Original Publications

Vol. 104 No. 5 November 2007

ORAL AND MAXILLOFACIAL PATHOLOGY Editor: Mark W. Lingen

Population-based study of treatment outcomes in squamous cellcarcinoma of the retromolar trigoneAbdulaziz Binahmed, BDS, MDent, MSc, FRCDC,a Richard W. Nason, MD, FRCSC,b

Ahmed A. Abdoh, PhD,c and George K. B. Sándor, MD, DDS, PhD, FRCDC, FRCSC, FACS,d

Winnipeg and Toronto, CanadaUNIVERSITY OF MANITOBA AND UNIVERSITY OF TORONTO

Background. Carcinoma of the retromolar trigone is relatively uncommon. High rates of local recurrence account fora relatively poor prognosis.Study design. A population-based historical cohort of 76 cases with biopsy-proven squamous cell carcinoma of the retromolartrigone were studied as a case series. Kaplan-Meier survival curves and log rank test were used for statistical analysis.Results. The mean age was 67.2 years. Fifty-six patients were male, 45% had T1 or T2 tumors, and 61% were stagedas N0. Treatment included radiotherapy in 35%, surgery alone in 26%, surgery and radiotherapy in 23%, and 16%received palliative treatment. The absolute and disease-specific survivals at 5 years were 51.4% and 67.7%, respectively.In patients treated with surgery, the resection margin status predicted the overall 5-year survival ( P � .027), with 75% ofpatients with negative margins surviving 5 years versus a survival of 0% of patients with involved margins.Conclusions. Squamous cell carcinoma of the retromolar trigone has a poor survival rate for early-stage disease. Adequate

surgical margins can improve survival. (Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007;104:662-5)

The retromolar trigone (RMT) is an ill-defined triangulararea in the oral cavity posterior to the upper and lowerthird molar teeth, with the maxillary tuberosity at its apex.The RMT is continuous laterally with the buccal mucosaand medially with the soft palate. Superiorly and inferiorlythe RMT blends with the mucoperiosteum of the maxil-

aConsultant, Department of Surgery, King Abdulaziz Medical City,Riyadh, Saudi Arabia; Former Fellow, Pediatric Oral and Maxillofa-cial Surgery, Hospital for Sick Children and University of Toronto,and Department of Surgical Oncology, Cancer Care Manitoba, Uni-versity of Manitoba.bAssociate Professor, Department of Surgical Oncology and Depart-ment of Surgery, Cancer Care Manitoba, University of Manitoba.cClinical Epidemiologist and Biostatistician, Department of Surgery,University of Manitoba.dClinical Director, Graduate Program in Oral and Maxillofacial Sur-gery and Anesthesia, Mount Sinai Hospital, Toronto; Coordinator,Pediatric Oral and Maxillofacial Surgery, Hospital for Sick Children andBloorview Kids Rehab, and Professor of Oral and Maxillofacial Surgery,University of Toronto; Docent, University of Oulu, Oulu, Finland.Received for publication Jan 5, 2007; returned for revision Mar 31,2007; accepted for publication Apr 6, 2007.1079-2104/$ - see front matter© 2007 Mosby, Inc. All rights reserved.

doi:10.1016/j.tripleo.2007.04.004

662

lary and mandibular alveoli.1 The RMT is a relativelyuncommon site for oral squamous cell carcinoma. Theetiology of these tumors, as for other sites of oral squa-mous cell carcinoma, has been linked to tobacco andalcohol use.2,3 Tumors of the RMT are often confused withtumors of the tonsil, soft palate, or alveolar ridge. Tumors inthis region are characterized by early invasion of the mandi-ble and the adjacent buccal mucosa and soft palate (Fig. 1).3,4

They are often diagnosed at an advanced stage owing theabsence of early symptoms. The prognosis tends to be poorbecause of the advanced stage of disease at the time ofpresentation.2,3,4 In their advanced stage, pain and trismus arecommon from the invasion of the buccal, lingual, and inferioralveolar branches of the mandibular nerve and the muscles ofmastication. Squamous cell carcinoma of the RMT has beentreated with primary surgery, primary radiation, and withcombinations of surgery, radiotherapy, and chemotherapywith varying success.3-6

The objective of the present population-based histor-ical case-series study was to evaluate outcomes in pre-viously untreated patients with squamous cell carci-noma of the RMT and to assess the impact of different

treatment modalities on recurrence and survival.

OOOOEVolume 104, Number 5 Binahmed et al. 663

MATERIALS AND METHODSThe charts of 88 patients registered with squamous cell

carcinoma of the RMT in the Province of Manitoba,Canada, from January 1975 to January 2004 were evalu-ated. Approval had been obtained to review these recordsfrom the University of Manitoba as part of a protocolentitled “A Quality Management Program for Head andNeck Oncology” (protocol reference no. H2005:071).

Twelve patients were excluded from the study. Thiswas because of incorrect site coding (n � 3), insuffi-

Fig. 1. A, Clinical appearance of squamous cell carcinoma ofthe retromolar trigone at presentation. B, Clinical appearance ofsquamous cell carcinoma of the retromolar trigone at biopsy.

cient clinical data (n � 6), or pathology report absent or

inconsistent with squamous cell carcinoma (n � 3).Seventy-six cases of biopsy-proven previously untreatedsquamous cell carcinoma of the RMT were identified.One patient presented with bilateral disease 10 years apart.This patient was considered as 2 distinct cases.

Tumor staging was done using clinical data recordedat the time of initial assessment of each patient accord-ing to the TNM (tumor, node, and metastasis) classifi-cation system of the International Union Against Can-cer, 2002.7 After appropriate merges and datatransformation, statistical analysis was done usingSPSS software (SPSS, Chicago, IL). Descriptive statis-tics were presented as mean � standard deviation (SD)for continuous variables and percentage (%) and sam-ple size (n) for categoric variables. The comparisonbetween the study groups was done for baseline char-acteristics using Pearson chi-squared, with continuitycorrection when appropriate, or Fisher exact test forcategoric variables and t test or Mann-Whitney U testfor continuous variables. Kaplan-Meier curves8 com-paring time-to-success (survival) probabilities betweenthe 2 study groups, using the log rank test, were used.

RESULTSThere were 56 men and 20 women, with a mean age

of 67.22 � 10.3 years. Tobacco use, alcohol use, or acombination of both was documented in 66 patients(86%; Table I). Diabetes, hypertension and ischemicheart disease were the most common comorbiditiesrecorded. Twenty-three patients had 34 second primarytumors, of which 62% were synchronous and 55%involved the upper aerodigestive tract.

Eight patients had stage I lesions, whereas 25, 12,and 29 patients had stage II, III, and IV disease respec-tively (Table II). Two patients were not staged.Twenty-eight patients (36%) had clinically positivenodes at the time of presentation.

Twelve patients were treated with a palliative intent.

Table I. Patient demographics, squamous cell carci-noma of retromolar trigoneMale, n (%) 56 (74%)Female, n (%) 20 (26%)Age at presentation, yrs 67.26 � 10.3Tobacco and/or alcohol use, n (%) 66 (86%)

Table II. Stage of squamous cell carcinoma of retro-molar trigone at diagnosis and survivalStage of disease At diagnosis, n 5-year survival

I 8 75%II 25 57%III 12 56%IV 29 38%

Surgery was used as the primary treatment modality in

OOOOE664 Binahmed et al. November 2007

20 patients (26%). Fifteen patients received postopera-tive radiotherapy. Radiotherapy was used with curativeintent in 29 patients (39.5%). Three of these patientshad a neck dissection for persistent disease.

Intraoral excision of the primary tumor was per-formed in 5 patients, whereas the remainder required acheek flap for exposure. Thirty-two patients underwentmandibular resection, either a marginal resection (n �10) or segmental mandibulectomy (n � 22). Primaryclosure was possible in 11 patients. Margins on perma-nent section were clear in 18 patients, close (within2 mm) in 8, and involved in 7. Radiation therapy as aprimary treatment modality or as adjunctive treatmentwas delivered by conventional fractionation at a doseranging from 4,350 to 6,600 cGy.

Twenty-nine patients were managed with neck dis-section alone. Ten patients were treated with neckdissection and adjunctive radiotherapy. Thirty-sevenpatients were treated with radiation alone.

Twenty-eight patients failed treatment, 18 (64%) at theprimary site, 3 (10.7%) in the neck, 2 (7%) at the primarysite and in the neck. Four patients had distant metastasis,and the site of failure was not documented in 1 patient.

The overall median survival time was greater than 60months, and 51.4% of the patients survived the 5-yearperiod. When disease-specific survival was considered,67.7% of the patients survived the 5-year period andmedian survival time was still above 60 months. Themedian disease-free survival was 60 months, and40.3% of the patients survived 5 years.

There was no statistically significant difference be-tween stage of disease and overall survival. Five-yearoverall survival rates were 75%, 57%, 56%, and 38% forstages I, II, III, and IV, respectively (P � .2230; Fig. 2).Five-year overall survival rates for patients with N0 andN� neck disease were 55% and 47.5%, respectively (P �.5712). The 5-year overall survival by treatment modalityfor patients treated with surgery alone was 64%, withsurgery and radiotherapy 57%, and 45% when radiationwas elected as a single treatment modality (P � .1260).The 5-year overall survival for patients with clear, close(within 2 mm), or involved surgical margins was 68%,83%, and 0.0%, respectively (P � .0270).

DISCUSSIONThis series of 76 patients with RMT squamous cell

carcinoma is comparable to most series of oral cancerin many respects, including age and gender distribution,etiologic factors, and incidence of second primaries.However, in contrast to other series this is a population-based study, describing events over a decade within awell defined geographic boundary, i.e., the Province of

Manitoba with a population of 1.5 million.

The authors have developed the clinical impressionthat patients with early-stage disease of the RMT failtreatment more frequently than observed with similar-staged lesions of the tongue and floor of mouth. Sur-vival of greater than 80% has been reported for early-staged lesions of the tongue and floor of mouth.9-13

The previously reported prognosis for squamous cellcarcinoma of the RMT is relatively poor, with overallsurvival ranging from 26% to 80%.3-6,14-17 Unlike theearlier studies, the overall survival of 51.4% in the presentseries does not reflect a selection bias for patients andinitial treatment modality, such as radical surgery3 ortreatment with radiation alone.5,14,15 The outcome re-ported in this study includes all patients managed specif-ically with squamous cell carcinoma of the RMT.

Clinical stage and survival did not show a significantcorrelation in the present review. The lack of signifi-cance may be a function of the small sample number ineach of the subdivided groups. Although Kowalskiet al.3 reported similar findings, other investigators15,16

have shown a significant influence on stage and sur-vival with both univariate and multivariate analysis.

The initial treatment modality did not affect survivalin the present series, which is again consistent withprevious reports.3-5 Surgery seems to have the bestsurvival rate, even if this is not statistically significant.The lack of success may have been due to the smallsample number in each treated group. More recentreports of treatment outcomes for squamous cell carci-noma of the RMT suggest that survival is improvedwith surgery followed by adjunctive radiotherapy.2,14,16

The surgical margin status impacted local-regionaltreatment failure, as reflected by a significant decrease inoverall survival with positive margins. This was the onlysignificant predictor of outcome in the present series. This

Fig. 2. Overall survival at 5 years by stage of disease atpresentation, using Kaplan-Meier analysis.

is consistent with some previous reports involving other

OOOOEVolume 104, Number 5 Binahmed et al. 665

sites and types of oral cancer.18,19 However, the marginstatus was not found to predict survival in some otherseries.3,14 This outcome was attributed to the use of post-operative radiotherapy by one investigator.14

The lack of concordance between studies with re-spect to prognostic factors for squamous cell carcinomaof the RMT again reflects small sample sizes withdifferences in patient selection. Larger multicenter pop-ulation-based studies using appropriate multivariatemodels are necessary to resolve these issues.

There is no agreement on the optimal managementof these tumors. Early reports20 recommended surgicaltreatment over radiotherapy. Kowalski et al.3 describedtheir experience with extended “commando” operationsin 114 patients with or without postoperative radiationand reported a 5-year absolute survival rate of 55.3%.They concluded that adjunctive radiotherapy improvedlocal and regional control.3 The experience reviewedby Byers et al.4 in 1984 showed a recurrence rate of16% in patients treated with radiation therapy alone and18% in patients treated with postoperative radiation.The 5-year absolute survival was 26%. In another re-port from the same institution, Lo et al.5 reviewedexperience with irradiation of the anterior faucial pillarand the RMT. The 5-year determinant survival was83%. In a more recent study Hao et al.17 reported a5-year actuarial survival of 60.6% in 50 patients treatedwith surgery, with 24 receiving adjuvant therapy.

Huang et al.,14 in their review of 65 patients, showeda 5-year disease-free survival of 31% with radiationalone, compared with 90% and 63% with surgery andpre- or postoperative radiation, respectively. Those au-thors recommended combined surgery and radiother-apy for all stages of disease. The most recent review byMendenhall et al.16 reported 5-year overall survivals of40% and 56% in 35 patients managed with radiationalone and 64 managed with surgery and radiotherapy,respectively. Using a multivariate analysis they foundthe stage (P � .0476) and the treatment modality (P �.0098) to be significant predictors of survival. Oneother study has shown improved survival with the ad-dition of chemotherapy to surgery and radiation.5

The optimal management of squamous cell carci-noma of the RMT cannot be determined from this orany other single review. When considering relativelyuncommon tumors such as squamous cell carcinoma ofthe RMT, there is a need for the head and neck oncol-ogy community to pool population-based outcome datato establish accurate prognostic indicators.

REFERENCES1. McGregor I, McGregor F, Cancer of the face and mouth. Pathol-

ogy and management for surgeons. Edinburgh: Churchill Living-

stone; 1986. p. 499-510.

2. Antoniadoes K, Lazardidis N, Vahtsevanos K, Hadjipetrou L,Antoniadoes V, Karakasis D. Treatment of squamous cell carci-noma of the anterior faucial pillar–retromolar trigone. Oral Oncol2003;39:680-6.

3. Kowalski L, Hashimoto I, Magrin J. End results of 114 extended“commando” operation for the retromolar trigone carcinoma.Am J Surg 1993;166:374-9.

4. Byers R, Anderson B, Schwars E. Treatment of squamous cellcarcinoma of the retromolar trigone. Am J Clin Oncol1984;7:647-52.

5. Lo K, Fletcher G, Byers R. Results of irradiation in the squamouscell carcinoma of the anterior facuial pillar–retromolar trigone.Int J Radiat Oncol Biol Phys 1987;13:969-74.

6. Genden E, Ferlito A, Shaha A, Rinaldo A. Management ofcancer of the retromolar trigone. Oral Oncol 2003;39:633-7.

7. International Union Against Cancer. TNM classification of ma-lignant tumors. 6th ed. New York: Wiley-Liss; 2002. p. 22-6.

8. Kaplan ER, Meier P. Non-parametric estimation from incom-plete observations. J Am Stat Assoc 1958;53:457-81.

9. Franceschi D, Gupta R, Spiro RH, Shah IP. Improved survival inthe treatment of squamous cell carcinoma of the tongue. Am JSurg 1993;166:360-5.

10. Lehman RH, Cox ID, Belson TP, Yale RS, Byhardt RW, ToohillRI, Malin TC. Recurrence patterns by treatment modality ofcarcinomas of the floor of the mouth and tongue. Am J Otolar-yngol 1982;3:174-81.

11. Shaha AR, Spiro RH, Shah IP, Strong EW. Squamous carcinomaof the floor of the mouth. Am J Surg 1984;148:455-9.

12. Hicks WL Jr, Loree TR, Garcia RI, Maamoun S, Marshall D,Omer JB, et al. Squamous cell carcinoma of the floor of themouth: a 20 year review. Head Neck 1997;19:400-5.

13. Woolgar JA, Rogers S, West CR, Errington RD, Brown JS,Vaughan ED. Survival and patterns of recurrence in 200 oralcancer patients treated by radical surgery and dissection. OralOncol 1999;35:257-65.

14. Huang C, Chao K, Tsai J, Simpson JR, Haughey B, Spector GJ,Sessions DG. Cancer of the retromolar trigone: long-term radi-ation therapy outcome. Head Neck Surg 2001;23:758-63.

15. Ayad T, Gelinas M, Guertin L, Larochelle D, Vecchio P, TabetJC, et al. Retromolar trigone carcinoma treated by primary radi-ation therapy: an alternative to the primary surgical approach.Arch Otolaryngol Head Neck Surg 2005;131:576-82.

16. Mendenhall WM, Morris CG, Amdur RJ, Werning JW, VillaretDB. Retromolar trigone squamous cell carcinoma treated withradiotherapy alone or combined with surgery. Cancer 2005;103:2320-2325.

17. Hao SP, Tsang NM, Chang KP, Chen CK, Huang SS. Treatmentof squamous cell carcinoma of the retromolar trigone. Laryngo-scope 2006;116:916-20.

18. Looser K, Shah J, Strong E. The significance of positive marginsin surgically resected epidermoid carcinoma. Head Neck Surg1978;1:107-11.

19. Batsakis JG. Surgical excision margins: a pathologist’s perspec-tive. Adv Anat Pathol 1999;6:140-8.

20. Barbosa J: Cancer of the retromolar trigone. Arch Otolaryngol1959;69:19-30.

Reprint requests:

Professor George K. B. SándorThe Hospital for Sick ChildrenS-525, 555 University AvenueToronto, OntarioCanada M5G 1X8

george.sandor@utoronto.ca

Treatment outcomes in squamous cell carcinoma of themaxillary alveolus and palate: a population-based study

Abdulaziz Binahmed, BDS, MDent, MSc, FRCDCa, Richard W. Nason, MD, FRCSCb,Ali Hussain, DDS, MDent, FRCDCc, Ahmed A. Abdoh, PhDd, andGeorge K.B. Sándor, MD, DDS, PhD, FRCDC, FRCSC, FACSe Riyadh, Saudi Arabia,Winnipeg and Toronto, Canada, and Tampere and Oulu, FinlandKING ABDULAZIZ MEDICAL CITY, UNIVERSITY OF MANITOBA, HOSPITAL FOR SICK CHILDREN,UNIVERSITY OF TORONTO, UNIVERSITY AND TAMPERE AND UNIVERSITY OF OULU

Objective. This population-based historical cohort study evaluates the treatment outcomes of primary squamous cellcarcinoma of the maxillary alveolus and hard palate.Methods. A historical cohort of 37 cases of previously untreated biopsy-proven squamous cell carcinoma of the upperjaw registered in the Province of Manitoba from January 1975 to January 2004 was analyzed.Results. The tumor epicenter involved the maxillary alveolus in 26 patients and the hard palate in 11 patients. The meanage of the study population was 72.8 years and 67% were women with a documented tobacco use rate of 50%. Forty-onepercent had stage I or II disease, 51% stage III or IV, and 8% could not be staged. Treatment included radiotherapy as asingle modality (13.5%), surgery (38%), surgery and radiotherapy (24%), and palliative treatment (24%). Local recurrencewas observed in 10 patients with 6 failing at the primary site. The absolute and disease-free survival at 5 years was 33%and 62% respectively. The 5-year disease-free survival was 82% for stage I and II and 48% for stage III and IV (P � .056).No patient treated with radiotherapy as a single treatment modality survived 5 years. Disease-free survival for patientstreated with surgery, and surgery � radiotherapy, was 69% and 73% at 5 years, respectively (P � .001).Conclusions. Squamous cell carcinoma of the maxillary alveolus and palate differs from other oral cancers in that thepatients are relatively older with a slight female predilection. Treatment with surgery, with or without radiotherapy,

appears to improve disease control. (Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;105:750-4)

aConsultant, Department of Surgery, King Abdulaziz Medical City,Riyadh, Saudi Arabia; Former Fellow, Pediatric Oral and MaxillofacialSurgery, The Hospital for Sick Children and University of Toronto,Toronto, Canada and Department of Surgical Oncology, Cancer CareManitoba, University of Manitoba, Winnipeg, Manitoba, Canada.bAssociate Professor, Department of Surgical Oncology, Cancer CareManitoba, Department of Surgery, University of Manitoba, Win-nipeg, Manitoba, Canada.cAssistant Consultant, Al-Amiri Hospital, Kuwait; Former Fellow,Department of Surgical Oncology, Cancer Care Manitoba, Universityof Manitoba, Winnipeg, Manitoba, Canada.dClinical Epidemiologist and Biostatistician, Department of Surgery,University of Manitoba, Winnipeg, Manitoba, Canada.eClinical Director, Graduate Program in Oral and MaxillofacialSurgery and Anesthesia, Mount Sinai Hospital, Toronto; Coordi-nator, Pediatric Oral and Maxillofacial Surgery, Hospital for SickChildren and Bloorview Kids Rehab and Professor of Oral andMaxillofacial Surgery, University of Toronto, Toronto, Ontario,Canada; Professor, Regea Institute of Regenerative Medicine,University of Tampere, Tampere, Finland; Dosent, University ofOulu, Oulu, Finland.Received for publication Apr 7, 2007; returned for revision Sep 6,2007; accepted for publication Sep 7, 2007.1079-2104/$ - see front matter© 2008 Mosby, Inc. All rights reserved.

doi:10.1016/j.tripleo.2007.09.001

750

Squamous cell carcinoma (SCC) of the hard palate andmaxillary alveolar ridge could be considered as 1 sitesince the 2 areas are adjacent anatomically and sharesimilar clinical presentations and management. Theseare relatively uncommon oral cancers, with a reportedincidence ranging from 0.5% to 2.0%.1,2 Many otherinvestigators have even considered SCC of the maxil-lary and mandibular alveolar ridges as similar enti-ties.1,3-5 However, the behavior and management ofSCC involving the maxillary and mandibular alveolarridges is different (Fig. 1, A-D), and therefore theinterpretation of their results if combined may be dif-ficult. Other publications have not only compared andcontrasted SCC of these subsites but have includedother tumors of non–squamous cell origin such as sal-ivary gland neoplasms.6,7 This study examines treat-ment outcomes in a population-based cohort of patientssolely with primary SCC limited to the maxillary alve-olus or hard palate.

MATERIALS AND METHODSThe records of 62 patients diagnosed with SCC of the

upper jaw and registered in the Province of Manitobafrom January 1975 to January 2004 were evaluated.

Approval had been obtained to review these records

ication

OOOOEVolume 105, Number 6 Binahmed et al 751

from the University of Manitoba as part of a protocolentitled “A Quality Management Program for Head andNeck Oncology” Protocol Reference Number: H2005:071. Patients were excluded for the following reasons:incorrect site coding (n � 6), insufficient clinical data(n � 2), other pathology (n � 16), and patients seen inconsultation only (n � 1). There were 37 cases ofbiopsy-proven and previously untreated SCC of themaxillary alveolus and hard palate. Staging was doneusing clinical data recorded at the time of the initialassessment of each patient according to the TNM (tu-mor, node, and metastasis) classification system of theInternational Union Against Cancer, 2002.8

After appropriate merges and data transformation,statistical analysis was performed using SPSS software(SPSS Inc., Chicago, IL). Descriptive statistics are pre-sented as mean � standard deviation (SD) for contin-

Fig 1. A, Stage I squamous cell carcinoma lesion affectingmaxillary alveolus resected en bloc with teeth. C, Maxillaryobdurate the defects with oral nasal or oral antral communicin pack and obturating any oral-nasal or oral-antral commun

uous variables, percentage (%), and sample size (n) for

categorical variables. The comparison between thestudy groups was done according to baseline charac-teristics using the Pearson chi-square test with contin-uous correction when appropriate, Fisher’s exact testfor categorical variables, and t tests or Mann-WhitneyU test for continuous variables. Kaplan-Meier9 curveswere used to compare time-to-success or survival prob-abilities between the different study groups using theLog-rank test.

RESULTSThe tumor epicenter was found to involve the max-

illary alveolar ridge in 26 patients and the hard palate inthe remaining 11 patients (Fig. 2). There were 12 malesand 25 females, with a mean age of 72.8 years (SD12.1). Tobacco, alcohol consumption, or a combinationof both was documented in 19 (50%) patients. One

spect of right anterior maxillary alveolus. B, Right anteriorions were managed with a stent fabricated preoperatively toimmediately after resection. D, Stent secured in situ holdings.

labial aresect

ations

patient was a reverse cigarette smoker. Diabetes, hy-

OOOOE752 Binahmed et al June 2008

pertension, and ischemic heart disease were the mostcommon comorbidities recorded. Eight of the patientshad a total of 11second primary tumors, of which 42%were synchronous and involved the upper aerodigestivetract. Fourteen patients (38%) had pain as the majorpresenting symptom. Six patients had stage I lesions, 9stage II, 4 stage III, and 15 patients stage IV. Threepatients could not be staged. Seven patients (19%) hadclinically positive nodes at the time of presentation(Table I).

Eight patients (21.6%) were treated with a solelypalliative intent. Surgery was used as the primarytreatment modality in 23 patients with 9 of thesepatients receiving postoperative radiotherapy. Radio-therapy was used as a single treatment modality in 5patients (13.5%). One patient was treated with radio-therapy and chemotherapy (Table II). In the 23 pa-tients treated with surgery a transoral excision wasused in 11 cases. In 11 patients a cheek flap wasdeemed necessary for adequate exposure. An opera-tive report was not identified in 1 patient. Margins onpermanent section were found to be clear in 16patients, close or within 2 mm in 3 of the patients,and involved in 4 patients. Radiation therapy,whether used as an adjunctive or as a primary treat-ment modality, was delivered by conventional frac-tionation at doses ranging from 4125 to 7000 CGy.Four of the 7 patients presenting with positive lymphnodes were treated with a curative intent. Two ofthese patients were treated with radiation alone. In

Fig 2. Diagramatic representation of the distribution of thelocation of the epicenters of the 37 cases of squamous cellcarcinoma of the maxillary alveolus and palate reviewed inthis case series. Maxillary alveolus represented by triangles;hard palate represented by circles.

the remaining 25 node-negative patients who were

treated with curative intent, the neck was treatedelectively in 2.

Ten of the 29 patients treated with curative intentfailed treatment with 6 failing at the primary site and 4failing in the neck. The absolute and disease-free sur-vival at 5 years was 33% and 62% respectively (Figs. 3and 4). The 5-year disease-free survival (Fig. 5) was82% for stage I and II and 48% for stage III and IV (P� .056). The initial treatment modality, whether sur-gery, radiotherapy, or a combination of both, signifi-cantly influenced the 5-year survival rate (P � .001).No patient treated with radiotherapy as a single treat-ment modality survived 5 years or longer. The 5-yeardisease- free survival for patients treated with surgeryalone, and a combination of surgery and radiotherapywas 69% and 73% respectively. Margin status did notaffect survival in any of the cases with involved mar-gins.

DISCUSSIONSquamous cell carcinoma of the maxillary alveolus

and hard palate is relatively uncommon in Westernsociety.10-12 Petruzzelli and Mayers13 noted that neo-plasias, both SCC and non-SCC, of the upper jawconstitute only 2% of all head and neck malignanciesand 10% of oral cancers. The lesions appear to beconsiderably more common in India, accounting for40% to 55% of the oropharyngeal cancers.14,15 Therelatively low numbers of these tumors is the mostlikely reason these lesions are often grouped and re-ported together with other sites such as the mandibularalveolus3,5 and soft palate1,6 or combined with salivary

Table I. Stage of disease at time of presentationStage of disease No. of patients*

Stage I 6Stage II 9Stage III 4Stage IV 15

*Three patients were not staged.

Table II. Treatment by stageStage of disease Surgery Radiotherapy Combination

Stage I 5 0 0Stage II 7 1 1Stage III 0 1 3Stage IV 2 1 6*Not Staged 0 2 0

Eight patients were treated with palliative intent.*One patient was treated with radiation and chemotherapy.

gland tumors.1,5 In contrast to most other reports, this

OOOOEVolume 105, Number 6 Binahmed et al 753

study examines treatment outcomes in a site that isanatomically specific and is pathologically limited tosquamous cell carcinoma of the maxillary alveolus andhard palate in a population-based historical cohort.

The anatomical characteristics of the maxillary alve-olar ridge and hard palate are similar as together theyform the roof of the oral cavity and the floor of the nasalcavity and maxillary sinus. Full-thickness resection ofthe structures involved by tumor at this site will resultin oronasal and oral-antral communication, a problemunique to this site. This unique anatomic relationshipmakes the radiographic work-up with computed tomog-raphy (CT) scans and magnetic resonance imaging

Fig 3. Kaplan-Meier survival curves showing 5-year overallsurvival.

Fig 4. Kaplan-Meier survival curves showing 5-year overalldisease-free survival.

(MRI) scans crucial to determine tumor extent and

stage. This radiographic visualization ultimately helpsdecide the staging, the extent of surgical resection, andthe use of other modalities of treatment.

Therefore, the clinical presentation, the surgicalmanagement, and reconstructive considerations includ-ing the management of oronasal and oral-antral com-munication with immediate obturation, are similar forthese sites (Fig. 1, A to D). From a clinical perspective,it is often difficult in advanced tumors to determine theexact epicenter (Fig. 2). The maxillary alveolus andhard palate, in contrast to the remainder of the oralcavity, seem to share a female predilection in this andother reports.16,17 In searching the literature further, theauthors were unable to find reasons for this femalepredilection as of yet. However, based on the forego-ing, the authors feel that it is most appropriate toconsider SCC of the upper alveolus and hard palate asa single entity.

The overall or absolute survival of 33% at 5 yearsreported in this study is relatively low. Reported sur-vival rates for these sites, which range from 21% to76%,1,5 are difficult to interpret, as they are oftengrouped to include other sites and other pathologicalentities such as salivary gland tumors. The Improveddisease-free survival at 5 years of 62% reported in thecurrent study reflects the influence of advanced age ofthe patients with significant comorbidities on overallsurvival. These considerations as well as the frequencyof stage IV disease (40%), account for the relativelylarge number of patients not amenable to treatment withcurative intent.

Margin status did not affect survival or outcome inany of the cases with involved margins. The reality isthat this is a study of only 37 cases in a rare site and theauthors did not feel comfortable drawing conclusions

Fig 5. Kaplan-Meier survival curves showing 5-year survivalrelated to stage of disease.

from such a small sample.

OOOOE754 Binahmed et al June 2008

The number of participants in this study is too smallto examine prognostic factors in detail. In this reportthere was a trend toward decreased survival with ad-vanced stage of disease as observed in other series.16,17

Yokoo et al.18 were unable to correlate stage of diseasewith prognosis. In their series, 80% of patients hadstage IV tumors. They proposed a new classificationsystem dependent upon the involvement of the maxil-lary sinus or nasal floor. The authors of the presentstudy observed improved survival in patients treatedwith surgery. Arguments in favor of one treatmentmodality over another in this, and the other seriesreviewed, are of limited value because of small samplesize in the studies.

There seems to be better prognosis for Stage III andIV lesions at this site compared to the late stages atother oropharyngeal sites. This trend was confirmed inother series.16-18 The reasons for this are unknown andnot apparent in the literature. The authors feel that thismay be because of differences in lymphatic drainage atthe maxillary alveolar and palatal sites when comparedto other oropharyngeal sites. This tumor is more prev-alent in India and these numbers may also representsome geographic differences potentially due to differ-ences in social habits such as reverse cigarette smoking.The authors feel that the numbers in the current studyare too small to be able to make further conclusions atthis time.

In conclusion, squamous cell carcinoma of the max-illary alveolar ridge and hard palate differs from otheroral cancers in several aspects. SCC tends to occur atthese sites in a relatively older population with a femalepredilection. A relatively high number of patients werenot amenable to treatment with curative intent. Treat-ment with surgery, with or without radiotherapy, ap-pears to improve disease control. A future prospectivemulticenter study with a larger cohort of patients mayhelp to answer questions regarding the prognosis oftreatment related to the specific stages of this disease.

REFERENCES1. Evan J, Shah J. Epidermoid carcinoma of the palate. Am J Surg

1981;142:4551-5.

2. New G, Hallberg O. The end results of the treatment of malig-nant tumors of the palate. Surg Gynecol Obstet 1941;73:520-4.

3. Gomez D, Faucher A, Picot V, Siberchoicot F, Renaud-Salis JL,Bussieres E, et al. Outcome of squamous cell carcinoma of thegingiva: a follow-up study of 83 cases. J Craniomaxillofac Surg2000;28(6):331-5.

4. Soo K, Spiro R, King W, Harvey W, Strong E. Squamous cellcarcinoma. Am J Surg 1988;156:281-5.

5. Cady B, Catlin D. Epidermoid carcinoma of the gum. A 20-yearsurvey. Cancer 1969;23:551-69.

6. Shibuya H, Horiuchi J, Suzuki S, Amagasa M, Mashima K. Oralcarcinoma of the upper jaw. Results of radiation treatment. ActaRadiol Oncol 1984;23(5):331-5.

7. Chung C, Johns M, Cantrell R, Constable W. Radiotherapy in themanagement of primary malignancy of the hard palate. Laryn-goscope 1980;90(4):576-84.

8. International Union Against Cancer. TNM classification of ma-lignant tumors. 6th ed. New York: Wiley-Liss; 2002. p. 22-6.

9. Kaplan ER, Meier P. Non-parametric estimation from incom-plete observations. J Am Stat Assoc 1958;53:457-81.

10. Bhansali S. Malignant tumors of the buccal cavity: a clinicalanalysis of 970 cases. Clin Radiol 1961;12:299-308.

11. Ong J. Cancer of the oral and pharyngeal cavities. J R Coll SurgEdinb 1970;15:250-67.

12. Krolls S, Hoffman S. Squamous cell carcinoma of the oral softtissues: a statistical analysis of 14,253 cases by age, sex, and raceof patients. J Am Dent Assoc 1976;92:571-4.

13. Petruzzelli G, Mayers E. Malignant neoplasms of the hard palateand upper alveolar ridge. Oncology 1994;8:43-8.

14. Ramulu C, Reddy C. Carcinoma of the hard palate and itsrelation to reverse smoking. Int Surg 1972;57:636-40.

15. Reddy C. Carcinoma of hard palate in India in relation to reversesmoking of chuttas. J Natl Cancer Inst 1974;53(3):615-9.

16. Yorozu A, Sykes A, Slevin N. Carcinoma of the hard palatetreated with radiotherapy: a retrospective review of 31 cases.Oral Oncol 2001;37(6):493-7.

17. Chung K, Rahman S, Lim M, Constable W. Squamous cellcarcinoma of the hard palate. Int J Radiation Oncology Biol Phys1979;5(2):191-6.

18. Yokoo S, Umeda M, Komatsubara H, Shibuya Y, Komri T.Evaluation of the T-classification of the upper gingival and hardpalate carcinomas—a proposition for new criterion of T4. OralOncol 2002;38(4):378-82.

Reprint requests:

George K.B. Sándor, MD, DDS, PhDThe Hospital for Sick ChildrenS-525, 555 University AvenueToronto, Ontario, Canada M5G 1X8

george.sandor@utoronto.ca

Relationship of the maxillary artery and lateral pterygoidmuscle in a caucasian sampleAli Hussain, BSC, DDS, FRCDC,a Abdulaziz Binahmed, BDS, MDent, MSc, FRCDC,b

Algernoun Karim, PhD,c and George K. B. Sándor, MD, DDS, PhD, FRCDC, FRCSC, FACS,d

Winnipeg and Toronto, CanadaUNIVERSITY OF MANITOBA AND UNIVERSITY OF TORONTO

Objective. Several studies have investigated the relationship between the maxillary artery and the lateral pterygoidmuscle, yet controversy still exists regarding this relationship. The purpose of this study was to re-examine thisrelationship in a sample of caucasian cadavers in Canada.Study design. All soft tissues were removed from the mandibles of 44 caucasian cadavers. Mandibular osteotomieswere performed to expose the infratemporal fossa and to dissect the maxillary artery bilaterally. Once accomplished,the relationship of the second part of the maxillary artery to the lower head of the lateral pterygoid muscle, and anyvariation in symmetry, was recorded.Results. In the majority of cases (30/44), the artery was found lateral to the lower head of the lateral pterygoid muscle(71% in men and 65% in women). The maxillary artery was found medial to the lower head of the lateral pterygoidmuscle in only 14 of the cases (29% in men and 35% in women). No variations in the course of the maxillary arterywere noted between the 2 sides and between both genders.Conclusions. This study shows that the lateral or superficial course of the maxillary artery relative to the lower head ofthe lateral pterygoid muscle is more prevalent than the medial or deep course. This is in agreement with the majorityof previously published results. There were no variations in the course of the artery between the 2 sides of the same

cadaver or between cadavers of both genders. (Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;105:32-6)

The topographic relations of the second part of thehuman maxillary artery in the infratemporal fossa havebeen studied since the late nineteenth century.1-7 Anumber of reports showed the course of the second partof human maxillary artery in the infratemporal fossa tobe variable.3,8 Understanding the anatomy of the hu-man maxillary artery and its branches in the infratem-poral fossa is a prerequisite for many surgical andclinical procedures. For instance, tumors involving the

aAssistant consultant, Al-Amiri Hospital, Kuwait; Former Fellow,Department of Surgical Oncology, Cancer Care Manitoba, Universityof Manitoba.bConsultant, Department of Surgery, King Abdulaziz Medical City,Riyadh, Saudi Arabia; Former Fellow, Pediatric Oral and Maxillofa-cial Surgery, The Hospital for Sick Children and University ofToronto and Department of Surgical Oncology, Cancer Care Mani-toba, University of Manitoba.cProfessor, Department of Oral Biology, University of Manitoba.dClinical Director, Graduate Program in Oral and Maxillofacial Sur-gery and Anesthesia, Mount Sinai Hospital; Coordinator, PediatricOral and Maxillofacial Surgery, Hospital for Sick Children andBloorview Kids Rehab; Professor of Oral and Maxillofacial Surgery,University of Toronto; and Dosent, University of Oulu, Oulu, Fin-land.Received for publication Nov 10, 2006; returned for revision Mar 13,2007; accepted for publication Apr 11, 2007.1079-2104/$ - see front matter© 2008 Mosby, Inc. All rights reserved.

doi:10.1016/j.tripleo.2007.04.010

32

cavernous sinus and the floor of the middle cranialfossa, such as meningiomas and trigeminal schwanno-mas, may extend into the infratemporal fossa. Othertumors originating in the nasopharynx such as angiofi-bromas, in the maxillary sinus such as carcinomas, andin the parotid gland such as adenoid cystic carcinomacan also invade the infratemporal fossa. In clinicaldentistry, mandibular block techniques, including theGow Gates and Vazirani-Akinosi, are commonly usedto achieve anesthesia of the mandibular nerve. Carefuladministration of the local anesthetic is necessary toavoid intra-arterial injection into the maxillary artery.2,9

Knowledge of the anatomy of the human maxillaryartery in the infratemporal fossa is important to den-tists, surgeons, and interventional radiologists.10

The maxillary artery arises as 1 of the 2 terminalbranches of the external carotid artery behind the neckof the mandibular condyle, deep to or within the sub-stance of the parotid gland.11,12 From its origin, thecourse of the maxillary artery is divided into 3 parts.The first or “mandibular” part runs nearly horizontaland deep to the neck of the mandibular condyle, andsuperficial to the sphenomandibular ligament during itscourse to the lower border of the lower head of lateralpterygoid muscle in the infratemporal fossa. The sec-ond or “pterygoid” part runs obliquely forward either

superficial or deep to the lower head of the lateral

OOOOEVolume 105, Number 1 Hussain et al. 33

pterygoid muscle. At this site, it is embedded into thepterygoid venous plexus. Finally, the third or “ptery-gopalatine” part disappears from view by diving be-tween the two heads of lateral pterygoid muscle toreach the pterygopalataine fossa via the pterygomaxil-lary fissure. There, it gives off several branches thataccompany the corresponding branches of the maxil-lary nerve.

The variable course of the second part of the maxil-lary artery has stimulated a considerable number ofreports.3,8 Lauber5 found only 16 arteries (8.5%) out of200 cases to be superficial to the lower head of thelateral pterygoid muscle. However, the superficial po-sition of the maxillary artery seemed to occur inabout 54%-70% of the cases in Caucasians3,4,6,7,13

and in 69% in Africans.4 Adachi1 stated that thesuperficial position of the second part of the maxil-lary artery occurred with highest frequency in theJapanese population and referred to this as a “racialdifference.” In support of this idea, Laskar et al.4

found the artery deep to the muscle in 46% of 147cases in Caucasians and in 31% of 61 cases inAfricans, and therefore they coined the term “racialblood vessel.” In the present study the authors haverevisited the relationship of the second part of themaxillary artery to the lateral pterygoid muscle in asample of the Caucasian population in Canada. Therelationship of maxillary artery to the mandibularnerve and its branches was not a subject of this study.

MATERIAL AND METHODSThe present study is based on investigations in 44

cadavers (21 men and 23 women) that were em-balmed and preserved in 2% formalin in the grossanatomy dissection laboratory at the University ofManitoba. All specimens were caucasian of knownage and gender.

The maxillary arteries were dissected bilaterallyusing a lateral infratemporal approach. The coveringsoft tissues, including the skin, fascial layers, andsuperficial portion of the parotid gland and masseterand temporalis muscles, were removed. The lateralsurface of the ramus of the mandible was exposedalong with the external carotid artery and the originof the maxillary artery. This approach also exposedthe mandibular condyles. Using an angled oscillatingsaw (Stryker, Kalamazoo, MI), the zygomatic archeswere removed and the mandibles were osteotomizedat the neck of the condyles and at the symphysismenti. The attachment of the lateral pterygoid mus-cle to the pterygoid fovea and the tempromandibularjoint (TMJ) was preserved. The mandible was thenremoved from its muscular attachments, including

the medial pterygoid, anterior belly of digastric,

mylohyoid, geniohyoid, and genioglossus muscles,without damaging any underlying tissues (Fig. 1).These procedures were also carried out on the con-tralateral side.

Careful dissection of the first and second parts ofmaxillary artery was performed so that the intimaterelation of the second part of maxillary artery to thelower head of lateral pterygoid could be visualized.The deep (Fig. 2) or superficial location (Fig. 1) of thesecond part of the maxillary artery to the lower head ofthe lateral pterygoid muscle was recorded in every

Fig. 1. The mandible has been removed, after severing thetendon of the temporalis muscle to expose the underlyingtissue. The second portion of the maxillary artery (MA) hasbeen dissected to reveal its relationship to the lower head ofthe lateral pterygoid muscle (LPTM). The superficial, orlateral, position of the second portion of the MA relative tothe lower head of the LPTM is seen in this specimen.

Fig. 2. This specimen shows the deep, or medial, position ofthe second part of the maxillary artery (MA) to the lower headof the lateral pterygoid muscle (LPTM).

specimen. This relationship and any variations in sym-

OOOOE34 Hussain et al. January 2008

metry were checked and documented. Photographswere taken using a digital camera (Coolpix 4500; Ni-kon, Tokyo, Japan). The information collected wasstatistically analyzed using the 2-tailed binomial testassuming a 50% chance that the maxillary artery will besuperficial or deep to the inferior head of the lateralpterygoid muscle.

RESULTSSuperficial (lateral) course of the maxillary arteryto the lower head of the lateral pterygoid muscle

The lateral or superficial position of the second partof the maxillary artery in relation to the lower head ofthe lateral pterygoid muscle was found in 68% of all thespecimens examined. In the male cadavers, the secondpart of the maxillary artery was found to be superficialin 71% of the total number of cases (P � .052). In thefemale cadavers, the artery was found to be superficialto the lower head of the lateral pterygoid muscle in 65%of the total number of cases (P � .117). There was nostatistically significant variation in the course of themaxillary artery between both sides of the same ca-daver and between genders (Figs. 3 and 4).

Deep (medial) course of the maxillary artery tothe lower head of lateral pterygoid muscle

The frequency of the deep (medial) course of thesecond part of the maxillary artery was relatively low.The maxillary artery was found to run deep to the lowerhead of the lateral pterygoid muscle in only 32% of allthe cases. In the male cadavers, the incidence was 29%,and in the female cadavers it was 35% (Figs. 3 and 4).Although there was a trend for female cadavers todisplay the deep course of the maxillary artery moreoften than males, this difference was not statistically

Fig. 3. The percentages of the superficial and deep positionsof the maxillary artery in 21 male specimens (P � .052).

significant. Similarly, there was no statistically signif-

icant variation in the deep course of maxillary arterybetween the two sides in the same specimen and be-tween genders.

DISCUSSIONUnderstanding the course of the maxillary artery in

the infratemporal fossa with its variation is of potentialinterest to many practitioners. For instance, tumorsarising or extending into the infratemporal fossa viaforamen ovale or foramen rotundum, such as juvenilenasopharyngeal angiofibromas, are commonly excisedby lateral cranial base approaches (Fisch type C)8,11 orsubtemporal/infratemporal approaches.13-17 In theseapproaches, the entire contents of the infratemporalfossa are exposed, including the maxillary artery whichmay be greatly susceptible to damage or sacrifice, de-pending on the extent of the surgery and the size of thetumor. Therefore, knowing the course of the maxillaryartery and its variations will aid in surgical planningand in minimizing intraoperative or postoperative hem-orrhage.

Many orthognathic and TMJ surgical procedures areperformed in close proximity to the maxillary artery inthe infratemporal fossa. Mandibular osteotomies, suchas intraoral vertical ramus osteotomy or bilateral sag-ittal split osteotomy, are commonly performed to cor-rect dentofacial deformities.18 Open TMJ proceduresusing the preauricular approach are relatively com-monly used. These surgical procedures or approacheshave been developed to improve safety to minimize orprevent damage to vital anatomic structures such as themaxillary artery. Avoidance of the maxillary artery orone of its branches is necessary to prevent extensiveintraoperative or postoperative hemorrhage.18,19

Fig. 4. The percentages of the superficial and deep positionsof the maxillary artery in 23 female specimens (P � .117).

In clinical dentistry, mandibular blocks,12 such as

OOOOEVolume 105, Number 1 Hussain et al. 35

Gow-Gates or Vazirani-Akinosi, are very commonlyused techniques to deliver local anesthetic agents be-fore many dental procedures. The site of injection isextremely close to the branches of the maxillary artery,which may result in undesirable consequences, such asintrarterial injection or hematoma.20 The clinician mustbe fully aware of the course of the maxillary artery andits variations to prevent the side effects from intra-arterial local anesthetic injection.

The course of the maxillary artery in relation to theinferior head of the lateral pterygoid muscle has been acontroversial point. As a result, a number of studieshave been conducted to investigate this mutual relation-ship. A review of the literature shows that many authorsfound the maxillary artery superficial to the lower headof lateral pterygoid muscle. This finding ranged be-tween 54% and 70%.4,6,7,9,21,22 Moreover, Adachi1

found a higher frequency of superficiality (93%) in theJapanese population, raising the possibility of racialdifferences. However, Lauber’s5 investigation showedonly 8.5% of the Caucasian population with a superfi-cial maxillary artery. In the present study, the relation-ship of maxillary artery to the lower head of the lateralpterygoid muscle was generally similar to that of otherstudies. On one hand, a trend was noted in the lateralcourse of the maxillary artery relative to the lower headof the lateral pterygoid muscle to be more common inmen (71%) than in women (65%), supporting other datain the literature.4,6,7,9,21,22 On the other hand, the medialcourse was less common in men (29%) than in women(35%) in the present study. However, the course of themaxillary artery in relation to the lower head of thelateral pterygoid was symmetrical in all of the presentspecimens. This is not in agreement with the report byPretterklieber et al.23 That study is the only one thatshows asymmetry to exist in the course of the maxillaryartery between the 2 sides in the same specimen. Thatfinding of Pretterklieber et al. seems questionable be-cause there is no other supporting work in the literature.

Many authors have attempted to explain the potentialfactors that determine the position of maxillary arteryrelative to the lateral pterygoid muscle in the infratem-poral fossa. However, craniometric analysis, includ-ing facial height, cranial breadth or length, and ce-phalic index, did not reveal any correlation betweenthe variable position of the maxillary artery and themeasured craniometric parameters.9,21-26 In anotherstudy by Lurje,27 the lateral course of the maxillaryartery was found to be more common in brachycephalicand mesocephalic individuals. However, that studylacked any statistical evaluation and was contradictedby other studies such as that of Laskar et al.4 Onetheory, the metabolic hypothesis,28 was proposed to

explain the potential factors involved in angiogenesis

and the vascularization of growing tissues. However, itfailed to explain the factors responsible for variation inthe position of the second part of the maxillary arteryrelative to the lateral pterygoid muscle.

In conclusion, the present study confirms some ofthe earlier findings concerning the more commonsuperficial position of the maxillary artery in relationto the lateral pterygoid muscle. However, the find-ings that there was a trend for the deeper course ofthe second part of the maxillary artery to be slightlymore common in women than in men and that therewas no evidence of asymmetry in the course of theartery within the 2 sides of the same specimen wereunique to this study. The mechanism responsible fordetermining the relative position of the artery to thelateral pterygoid muscle is still unknown. Thepresent study helps to begin to define the question ofthe effect of ethnicity on the relative position of themaxillary artery to the inferior head of the lateralpterygoid muscle by studying a Caucasian popula-tion from Canada. Comparison with future studies inother ethnic groups19 will further help to answer thisquestion.

The relationship between the maxillary artery and thelateral pterygoid muscle may be important during majorfacial reconstruction, cleft craniofacial reconstruction,cancer surgery, and surgery to the TMJ. The anatomyof the terminal end of the maxillary artery should bekept in mind during surgical dissection.

REFERENCES1. Adachi B. Das Arteriensysteem der Japaner, vol. 1, Kyoto:

Maruzen; 1928. p. 90-112.2. Biermann H. Die chirurgische Bedeutung der Lagevariationen

der Arteria Maxillaris. Anat Anz 1943;94:289-309.3. Krizan Z. Beitrage zur deskriptiven und topographischen Anato-

mie der A. maxillaries. Acta Anat 1960;41:319-33.4. Laskar GW, Opdyke DL, Miller H. The position of the internal

maxillary artery and its questionable relation to the cephalicindex. Anat Rec 1951;109:119-26.

5. Lauber H. Uber einige Varietaten im Verlauf der Arteria maxil-laries interna. Anat Anz 1901;19:444-8.

6. Lurje A. On topographical anatomy of the inner maxillary artery.Acta Anat 1946;2:219-31.

7. Thompson A. Report of the committee of collective investigationof the Anatomical Society of Great Britain and Ireland for theyear 1889-1890. J Anat 1891;25:89-101.

8. Fisch U, Pillsbury HC. Infratemporal fossa approach to lesions inthe temporal bone and base of the skull. Arch Otolaryngol1997;105:99-107.

9. Skopakoff C. Uber die variabilitat im verlauf der arteria maxil-laries. Anat Anz 1968;123:531-46.

10. Lasjaunias P, Berenstein A. Surgical neuroangiography. 1. Func-tional anatomy of craniofacial arteries. New York: Springer; 1987.p. 85.

11. Allen WE, Kierl EL, Rothmann SLG. The maxillary artery incraniofacial pathology. Am J Roentgenol 1974;121:124-38.

12. Coleman RD, Smith RA. The anatomy of mandibular anesthesia:

OOOOE36 Hussain et al. January 2008

Review and analysis. Oral Surg Oral Med Oral Path Oral RadiolEndod 1982;54:148-52.

13. Fisch U. Infratemporal fossa approach to tumors of the temporalbone and base of the skull. J Laryngol Otol 1978;92:949-67.

14. Fisch U. Infratemporal fossa approach for glomus tumors of thetemporal bone. Ann Otol Rhinol Laryngol 1982;91:474-9.

15. Gates GA. The lateral facial approach to the nasopharynx andinfratemporal fossa. Otolaryngol Head Neck Surg 1988;99:321-5.

16. Holliday MJ. Lateral transtemporal-sphenoid approach to theskull base. Ear Nose Throat 1986;65:153-62.

17. Mickey B, Close L, Schaefer S, Samson D. A combined fronto-temporal and lateral infratemporal fossa approach to the skullbase, J Neurosurg 1988;68:678-83.

18. Fujimura K, Segami N, Kobayahi S. Antomical study of thecomplications of intraoral vertico-sagittal ramus osteotomy.J Oral Maxillofac Surg 2006;64:384-9.

19. Apinhsmit W, Methathrathip D, Ploytubtim S, Chompoopong,Ariyawatkul T, Lertsirithong A. Anatomical study of the maxil-lary artery at the pterygomaxillary fissure in a Thai population;its relationship to maxillary osteotomy. J Med Assoc Thailand2004;87:1212-7.

20. Frangiskos F, Stavrou E, Tsitsogianis H, Vardas E, Antonopou-lou I. Incidence of penetration of a blood vessel during inferioralveolar nerve block. Br J Oral Maxillofac Surg 2003;41:188-9.

21. Vrionis FD, Cano WG, Heilman CB. Microsurgical anatomy ofthe infratemporal fossa as viewed laterally and superiorly. Neu-

rosurgery 1996;39:777-85.

22. Orbay H, Unlu RE, Sensoz O, Comert A. Vascular anatomy ofthe lateral pterygoid muscle. Ann Plast Surg 2006;56:584-5.

23. Pretterklieber ML, Skopakoff C, Mayr R. The human maxillaryartery reinvestigated: topographical relations in the infratemporalfossa. Acta Anat 1991;142:281-7.

24. Macho GA. Cephalometric and craniometric age changes in adulthuman. Ann Human Biol 1986;13:49-61.

25. Sjovold T. Geschlechtsdiagnose am Sklett. In: Knussmann R,editor. Anthropologie. Handbuch der vergleichenden Biologiedes Menschen, vol. 1. Stuttgart: Fischer; 1988. p. 444-79.

26. Vertcauteeren M. Age effects and secular trend in a cross-sectional sample: applications to forehead dimensions in Belgianadults. Hum Biol 1990;62:681-8.

27. Lurje A. On topographical anatomy of the inner maxillary artery.Acta Anat 1946;2:219-31.

28. Adair TH, Gay WJ, Montani JP. Growth regulation of the vas-cular system: evidence for metabolic hypothesis. Am J Physiol1990;259:R393-404.

Reprint requests:

Professor George K. B. SándorThe Hospital for Sick ChildrenS-525, 555 University AvenueToronto, OntarioCanada M5G 1X8

george.sandor@utoronto.ca

Int. J. Oral Maxillofac. Surg. 2007; 36: 712–715doi:10.1016/j.ijom.2007.02.011, available online at http://www.sciencedirect.com

Clinical PaperHead and Neck Oncology

Clinical observations of theanatomy and function of themarginal mandibular nerve§

R. W. Nason, A. Binahmed, M. G. Torchia, J. Thliversis: Clinical observations of theanatomy and function of the marginal mandibular nerve. Int. J. Oral Maxillofac.Surg. 2007; 36: 712–715. # 2007 International Association of Oral and MaxillofacialSurgeons. Published by Elsevier Ltd. All rights reserved.

§ Presented at the 6th International Con-ference on Head and Neck Cancer, Washing-ton, August 2004.

0901-5027/080712 + 04 $30.00/0 # 2007 Interna

tional Association of Oral and Maxillofacial Surgeon

R. W. Nason1, A. Binahmed2,M. G. Torchia3, J. Thliversis4

1Department of Surgical Oncology, CancerCare Manitoba, Winnipeg, Manitoba, Canada;2Department of Surgery, King AbdulazizMedical City Riyadh, Saudi Arabia;3Department of Surgery, University ofManitoba, Winnipeg, Manitoba, Canada;4Department of Human Anatomy and CellScience, University of Manitoba, Winnipeg,Manitoba, Canada

Abstract. The objective of this study was to assess the anatomical variation of themarginal mandibular nerve, and evaluate the risk of nerve malfunction after neckdissection. The method involved clinical assessment of the anatomy and function ofthe marginal mandibular nerve in 133 neck dissections. When the neck wasextended the nerve was displaced in an anterior and downward direction with thelowest point 1.25 � 0.7 cm below the mandible between the posterior and anteriorfacial veins. The nerve was >1 cm below the lower border of the mandible in 54%of dissections. When the intent was to preserve the nerve, dysfunction was observedin 16 of 101 dissections (16%). The incidence of marginal mandibular nervedysfunction following neck dissection is comparable to that observed followingsubmandibular gland excision for benign disease. Placement of incisions 2 cmbelow the lower border of the mandible will put the nerve at risk in a significantnumber of patients.

Key words: marginal mandibular nerve; neckdissection.

Accepted for publication 9 February 2007Available online 27 March 2007

Injury to the marginal mandibular nerveresults in a significant cosmetic deformity.The deformity is caused by interruption ofnerve fibers to the depresssor anguli orisand the depressor labii inferioris thatresults in flattening and inversion of theipsilateral lower lip and the inability tomove the lip downward and laterally. Thedefect may not be apparent with the face inrepose. Elevation of the affected lower lipis evident when the patient smiles, and the

asymmetry is most obvious when thepatient cries – the ‘asymmetric cryingfacies’5,11. The reported incidence of mar-ginal mandibular nerve injury ranges from0 to 20% following submandibular glandremoval3,4,9. Injury to the marginal man-dibular nerve in the course of a neckdissection has been considered an accep-table outcome3,5 and the incidence ofnerve dysfunction in this context is notreported.

Knowledge of the course and anatomicrelations of the marginal mandibular nervein the upper neck is important in avoidinginjury. The nerve is usually described as it

relates to the lower border of the mandibleand the facial vessels. The seminal workon the surgical anatomy of the marginalmandibular nerve is based on DINGMAN &GRABB’s study2 of 100 embalmed cadaverfacial halves. They described two or morerami of this branch of the facial nerve atthe level of the angle of the mandible. In98% of specimens the nerve passed down-ward and forward over the surface of theposterior facial vein and more anteriorlythe nerve crossed the anterior surface ofthe anterior facial vein in 100% of speci-mens. Immediately anterior to the facialvein the branches of the mandibular ramus

s. Published by Elsevier Ltd. All rights reserved.

Marginal mandibular nerve 713

Table 1. Management of the marginal mandibular nerve in 133 neck dissections

Management intent Nerve visualized Nerve not visualized Total

Preserve 76 (68) 36 (32) 112Sacrifice 9 (43) 12 (57) 21

Total 85 (64) 48 (36) 133

Numbers in parentheses are percentages.

passed superficial, deep or on both sur-faces of the facial artery. They tended tobe superficial to the artery. It was foundthat posterior to the facial artery the man-dibular ramus ran above the inferior bor-der of the mandible in 81% of specimensand in the other 19% the nerve or one ormore of its branches ran in an arc 1 cm orless below the inferior border of the mand-ible. Anterior to the facial artery allbranches were above the inferior borderof the mandible. DINGMAN & GRABB’s2

observations and other cadaver dissec-tions12 imply that incisions placed 2 cmbelow the lower border of the mandiblewill avoid injury to the nerve in allinstances. Clinical observations1,6 suggestthat the nerve is consistently below theinferior border of the mandible. BAKER &CONLEY

1 noted that in individuals with laxand atrophic tissue the nerve could be aslow as 3–4 cm below this point.

This study reports clinical observationson the surgical relations of the marginalmandibular nerve at the time of neckdissection. The incidence of nerve dys-function following neck dissection isdocumented.

Fig. 1. The relation of the marginal mandibular nerve (MMN) to the lower border of themandible (MLB), anterior facial vein (AFV) and posterior facial vein (PFV) with the neck in aneutral position (A) and in an extended position (B) with downward traction on the cervicalfascia.

Methods

Clinical observations of the anatomy andfunction of the marginal mandibular nervewere recorded in 133 neck dissections in121 patients. There were 78 males and 48females with a mean age of 54.31 � 17.0years. The neck dissections were done forsquamous cell carcinoma (n = 93), thyroidcancer (n = 21), salivary gland malig-nancy (n = 9), melanoma (n = 8), angio-sarcoma (n = 1), and Levoise-Bensaudsyndrome (n = 1). The neck dissectionswere classified as comprehensive in 72necks, selective in 59 and extended in 2.Level I nodes were removed in 109 (83%)of the dissections. Twenty-five patientshad received radiotherapy prior to theneck dissection.

All patients were positioned with a rollunder the shoulders to maintain extensionof the neck. Skin incisions in the upperneck were placed in skin creases 2–4 cmbelow the lower border of the mandible. Ingeneral, skin flaps were elevated withsharp dissection in a sub-platysmal plane.When Level I was dissected an attemptwas made to identify the marginal man-dibular nerve using standard land-marks2,6,10. This generally involvedobservation of the nerve coursing super-ficial to the anterior facial vein. The dis-tance from the lower border of themandible to the lowest point of the nervein the neck was measured. The relation-

ship of the lowest point of the marginalmandibular nerve to the anterior facialvein and posterior facial vein (anteriordivision of retromandibular vein) wasnoted. These measurements and relation-ships were recorded with the head in aneutral position, and in an extended posi-tion with downward traction on the invest-ing layer of the deep cervical fascia.

The functional status of the marginalmandibular nerve was recorded at the timeof each postoperative visit. The patientswere generally followed at intervals of 2–3months. The average length of follow upwas 661.18 days (range: 7–1832 days;median: 541 days). Statistical analysiswas performed with a Student’s t-test,Chi-square, or multivariate analysis whereappropriate.

Results

The marginal mandibular nerve was man-aged with the intent to preserve function in112 (84%) necks. The nerve was visua-lized in 76 (68%) of these patients. Adecision to sacrifice the nerve, based ononcological considerations, was made in21 dissections (Table 1).

Observations on the course and surgicalrelations of the marginal mandibular nervewere recorded in 85 patients. With thehead in neutral position the nerve was atits lowest point 0.31 � 0.4 cm (range 0.0–1.5 cm) below the inferior border of themandible in the region of the posteriorfacial vein (Fig. 1). With the neckextended the nerve was displaced in ananterior and downward direction with the

714 Nason et al.

Table 2. Distance marginal mandibular nerve identified below the mandible with the neck inneutral and extended position in 78 neck dissections

Distance below mandible (cm) Neutral Extended

>1 3 (4) 42 (54)>2 0 (0) 8 (10)

Numbers in parentheses are percentages.

lowest point 1.25 � 0.7 cm (range 0–3.0 cm) below the mandible between theposterior and anterior facial veins(P < 0.01). In the extended position thenerve was >1 cm below the lower borderof the mandible in 42 (54%) and >2 cm in8 (10%) patients (Table 2).

The functional status of the nerve wasrecorded following 121 neck dissections.Praxia was recorded in 18 (90%) patientswhere the nerve was visualized and delib-erately divided (n = 9) or presumeddivided with dissection in a plane super-ficial to the platysma (n = 11) for oncolo-gic reasons. In the remaining 101 patientswhere the intent was to preserve the nerve,praxia was observed in the immediatepostoperative period in 29 and persistedin 16 (16%). Temporary dysfunction ofthe marginal mandibular nerve had usuallyresolved by the second or third postopera-tive visits (3–6 months). Praxia was notobserved in the 24 patients where level Ilymph nodes were not dissected. The inci-dence in the 77 patients where a level Idissection was performed with the intent

Table 3. Relationship between clinical and tremarginal mandibular nerve dysfunction in 77 patito preserve nerve function

Variable Praxia

GenderMale 11 (19.6)Female 5 (23.8)

Age>57 11 (24.4)�57 5 (15.6)

PathologySCC 12 (18.5)Other 4 (33.3)

Preoperative radiotherapyYes 7 (36.8)No 9 (15.5)

Oral/oropharyngeal resectionYes 4 (11.1)No 12 (29.3)

Level I nodes positiveYes 1 (10.0)No 15 (22.4)

Nerve visualizedYes 13 (20.0)No 3 (25.0)

Numbers in parentheses are percentages. SCC,

to preserve nerve function was 16/77(21%). The relationship between nervedysfunction and clinical and treatment-related factors in this subset is shown inTable 3. The incidence of nerve dysfunc-tion appeared to be higher when the neckdissection followed radiotherapy. Thisapproached, but did not achieve, statisticalsignificance. Visualization of the nerveduring neck dissection did not affect thefunctional outcome. The use of multivari-ate models did not establish any indepen-dent predictors of nerve dysfunction.

Discussion

The complication of marginal mandibularnerve injury following neck dissectionsfor malignant disease has not receivedmuch attention. This is not surprising ashistorically other morbidities of majorhead and neck resections have over-whelmed this disability. The significanceof asymmetry of the lower lip increases asmorbidity from other aspects of the treat-ment is minimized through multimodal

atment-related factors and the incidence ofents where level I was dissected with the intent

Normal function P-value

45 0.75516

34 0.40527

53 0.2598

12 0.05849

32 0.08929

9 0.67852

52 0.7069

squamous cell carcinoma.

treatment plans that incorporate conserva-tive surgical approaches and organ pre-servation protocols. The frequency ofnerve dysfunction observed in this pro-spective evaluation of neck dissections, iscomparable to that reported in retrospec-tive reviews of outcome following exci-sion of the submandibular gland forbenign disease. It is recognized that retro-spective studies underestimate the trueincidence of complications.

This series represents a heterogeneouspopulation of patients with respect topathological, clinical and treatment-related parameters including the extentof neck dissection. In the cases wherepraxia of the marginal mandibular nervewas observed following a neck dissectionwith visualized or presumed division ofthe nerve, the outcome of the nerve couldusually be predicted preoperatively. Thesepatients generally had advanced local orregional disease. Patients where level Inodes were not dissected were includedin the functional assessment as the prox-imal marginal mandibular nerve isexposed to potential traction type injuryfrom retraction to expose high jugulodi-gastric lymph nodes. The absence of nervedysfunction in this cohort of patients maybe explained by the fact that the nerve iswell protected by surrounding soft tissue.When level I was dissected and the intentwas to preserve the nerve, the incidence ofpraxia appeared to be higher after preo-perative radiotherapy. The majority ofpatients in this series were managed sur-gically within 6–8 weeks of completion ofradiotherapy. Tissue planes tended to beedematous, but the anatomy was not dis-torted. The fact that no independent pre-dictors of nerve dysfunction wereidentified in patients where level I wasdissected suggests that anatomic consid-erations are paramount in preserving thefunction of the marginal mandibularnerve.

The anatomy of the marginal mandib-ular nerve in its vulnerable position in theupper neck is variable between patients.The course and position of the nerve alsovaries with the position of the head anddownward traction on the investing layerof cervical fascia. The dynamic nature ofthe nerve was evident in this study. Thisexplains, in part, the discrepanciesbetween cadaver studies and clinicalobservations. Cadaver tissue is contractedand relatively immobile. Of interest, whenfresh as opposed to embalmed cadavermaterial is studied, the nerve is consis-tently reported below the lower border ofthe mandible. In a series of 22 fresh cada-vers SAVARY et al.7 observed the nerve

Marginal mandibular nerve 715

below the mandible in 63% of specimensposterior to the facial artery and in 27%anterior to the artery. They observed somebranches as low as 3–4 cm below thelower border of the mandible. NELSON &GINGRASS

6 noted that the marginal man-dibular nerve was well below the inferiorborder of the mandible in almost everyinstance in fresh cadaver specimens andclinical dissections. Both these reportsrecommended cervical incisions severalcentimetres below the inferior border ofthe mandible. The present study confirmsthe consistent course of the marginal nervebelow the inferior border of the mandible.The observation by BAKER & CONLEY

1 thatthe nerve is drawn downward when theneck is extended is confirmed. The place-ment of cervical incisions 2 cm below thelower border of mandible will place thenerve at risk in a significant number ofpatients.

Detailed anatomic dissections2,6,8 showthat there are often multiple contributionsto the marginal mandibular nerve, and thatit can be difficult to discern between man-dibular and platysmal branches. DINGMAN

& GRABB2 note that in many specimens

fine branches could be seen running alongthe lower border of the mandible, some upto 2 cm below it. They stated that thesebranches terminated in and innervated theplatysma in all specimens. NELSON & GIN-

GRASS6 argued that these same branches,

when followed anteriorly in fresh cadaverspecimens and in clinical practice,ascended over the lower border of themandible to innervate specific lip depres-sors. They contend that these branchesshould be identified as mandibular andnot cervical branches. SKANDALAKIS

et al.8 describes an anterior ramus of thecervical division of the facial nerve thatjoins the mandibular ramus to contributeto the innervation of the lower lip. Thereare also observations that the platysma,innervated by the cervical division of thefacial nerve, contributes to depression ofthe lower lip11. Division of the cervicalbranch of the facial nerve or the platysmacan result in a pseudo-paralysis of themarginal mandibular nerve that usuallyrecovers spontaneously. Detailed dissec-tions of the marginal mandibular nerve

were not done in the present study. Froma practical point of view there is usually asingle identifiable nerve ramus coursingsuperficial to the facial veins that results inasymmetry of the lower lip if divided.Stern10 indicates that if the nerve doesnot course above the inferior border ofthe mandible within 2 cm of the facialvessels it is not the marginal mandibularnerve.

A low cervical incision with a ‘non-visualization’ approach to the marginalmandibular nerve is supported4,9 for man-agement of benign disease of the subman-dibular gland. In this study visualization ofthe nerve during the course of a neckdissection was not associated with a lowerincidence of nerve dysfunction. The fol-lowing technical considerations shouldminimize injury to the marginal mandib-ular nerve in the context of a neck dissec-tion. Irrespective of the site of the skinincision, skin flaps should be carefullyelevated in a plane immediately deep tothe platysma and superficial to the invest-ing layer of deep cervical fascia and mar-ginal mandibular nerve. It is not the levelof the skin incision that is important butthe level of transection of the investinglayer of cervical fascia. Knowledge of thedynamic and variable location of the mar-ginal mandibular nerve relative to theinferior border of the mandible is usefulin this regard. The decision to visualize thenerve needs to be individualized. If it isdeemed necessary to divide the cervicalfascia within 3–4 cm of the inferior borderof the mandible it is the authors’ opinionthat an attempt should be made to visua-lize the course of the nerve. The nerve isthen preserved or sacrificed based on theoncologic objectives of the procedure.Electrical cautery should be used spar-ingly, and attention should be directedto avoiding traction or pressure injuryfrom retractors.

References

1. Baker DC, Conley J. Avoiding facialnerve injuries in rhytidectomy. PlastReconstr Surg 1979: 64: 781–795.

2. Dingman RO, Grabb WC. Surgical anat-omy of the mandibular ramus of the facial

nerve based on the dissection of 100facial halves. Plast Reconstr Surg 1962:29: 266–272.

3. Hald J, Andreassen UK. Submandibu-lar gland excision: short- and long-termcomplications. J Otorhinolaryngol RelatSpec 1994: 56(2):87–91.

4. Ichimura K, Nibu K-I, Tanaka T.Nerve paralysis after surgery in the sub-mandibular triangle: review of Universityof Tokyo Hospital experience. Head Neck1997: 19: 48–53.

5. Moffat DA, Ramsden RT. The defor-mity produced by a palsy of the marginalmandibular branch of the facial nerve. JLaryngol Otol 1977: 91: 401–406.

6. Nelson DW, Gingrass RP. Anatomy ofthe mandibular branches of the facialnerve. Plast Reconstr Surg 1979: 64:479–482.

7. Savary V, Robert R, Rogez JM, Arm-

strong O, Leborgne J. The mandibularmarginal ramus of the facial nerve: ananatomic and clinical study. Surg RadiolAnat 1997: 19: 69–72.

8. Skandalakis JE, Gray SW, Rowe JS.Surgical anatomy of the submandibulartriangle. Am Surg 1979: 590–596.

9. Smith WP, Peters WJN, Markus AF.Submandibular gland surgery: an audit ofclinical findings, pathology and post-operative morbidity. Ann Royal Coll SurgEngl 1993: 75: 164–167.

10. Stern S. Anatomic correlates of headand neck surgery. Head Neck 1992: 14:328–331.

11. Tulley P, Webb A, Chana JS, Tan T,Hudson D, Grobbelaar AO, Harrison

DH. Paralysis of the marginal mandibularbranch of the facial nerve: treatmentoptions. Br J Plast Surg 2000: 53: 378–385.

12. Ziarah HA, Atkinson ME. The surgicalanatomy of the mandibular distribution ofthe facial nerve. Br J Oral Surg 1981: 19:159–170.

Address:Richard W. NasonDepartment of SurgeryUniversity of ManitobaHealth Sciences Centre820 Sherbrook StreetWinnipegManitoba R3A 1R9CanadaTel: +1 204 787 1094Fax: +1 204 787 4837E-mail: nasonrw@cc.umanitoba.ca

Oral Oncology (2007) 43, 780–784

ava i lab le at www.sc iencedi rec t . com

journal homepage: ht tp : / / in t l .e lsevierheal th .com/ journals /oron/

The clinical significance of the positivesurgical margin in oral cancer q

Abdulaziz Binahmed a, Richard W. Nason b,*, Ahmed A. Abdoh c

a Department of Oral and Maxillofacial Surgery, Winnipeg, Man., Canadab Department of Surgical Oncology, Cancer Care Manitoba, Winnipeg, Man., Canadac Department of Surgery, University of Manitoba, Winnipeg, Man., Canada

Received 1 August 2006; received in revised form 2 October 2006; accepted 4 October 2006Available online 14 December 2006

Summary The objective of surgical management of squamous cell carcinoma of the oral cav-ity is adequate resection with a clear margin. This study examines the significance of the posi-tive surgical margin. An historical cohort of 425 patients from the cancer registry of theProvince of Manitoba with squamous cell carcinoma of the oral cavity treated with surgery±radiotherapy was examined. A Cox’s proportional hazard model was used to examine the inde-pendent effect of surgical margins on five-year survival. Seventy-two percent of tumorsinvolved the tongue and floor of mouth, and 43% of patients presented with Stage III and IV dis-ease. The 5-year absolute and disease specific survivals were 62% and 74.5% respectively. Sur-vival was related to age >65 years (P = 0.0177), T-Stage (P = 0.0002), and N-Stage (P = 0.0465).Patients with clear margins had a survival rate of 69% at 5 yrs (median survival >60 mos) com-pared to 58% with close (median survival >60 mos) and 38% with involved margins (median sur-vival 31 mos, P = .0000). After controlling for significant prognostic factors, involved surgicalmargins increased the risk of death at 5 years by 90% (HR 1.9, 95% CI 1.2,2.9, P = 0.0026).The status of the surgical margin is an important predictor of outcome. The surgical margin,in contrast to the other prognostic indicators, is under the direct control of the surgeon.

�c 2006 Elsevier Ltd. All rights reserved.

KEYWORDSOral cancer;Treatment outcome;Surgical margins

1d

M

sW2

368-8375/$ - see front matter �c 2006 Elsevier Ltd. All rights reservedoi:10.1016/j.oraloncology.2006.10.001

q Presented at the 17th International Conference on Oral andaxillofacial Surgery. August 2005, Vienna, Austria.* Corresponding author. Address: Department of Surgery, Univer-ity of Manitoba, Health Sciences Centre, 820 Sherbrook St.,innipeg, Mani., Canada R3A 1R9. Tel.: +1 204 787 1094; fax: +104 787 4837.E-mail address: nasonrw@cc.umanitoba.ca (R.W. Nason).

Introduction

Microscopic tumor at, or close to the inked resection marginincreases local recurrence by a factor of two or more inmost series.1–10 A negative impact on survival is not asclear, accepted by some2,3,6,11,12 but not other investiga-tors.4,5,9,13,14 The status of the surgical margin has beenassociated with several factors that influence survival in

.

Table 1 Relationship of the status of the surgical margin toclinical data

Variable Involved(n = 62)

Close(n = 68)

Clear(n = 295)

Agea

>65 (n = 199) 33(16.6) 32(16.1) 134(67.3)665 (n = 225) 29(12.9) 36(16.0) 160(71.1)

GenderMale (n = 250) 41(16.4) 36(14.4) 173(69.2)Female (n = 175) 21(12.0) 32(18.3) 122(69.7)

SiteFloor of mouth(n = 169)

29(17.2) 26(15.4) 114(67.4)

Tongue (n = 129) 12(9.3) 17(13.2) 100(77.5)Other (n = 127) 21(16.5) 25(19.7) 81(63.8)

T-statusT1/2 (n = 297) 33(11.1) 46(15.5) 218(73.4)T3/4 (n = 119) 27(22.7) 19(16.0) 73(61.3)

N-statusN0 (308) 34(11.0) 48(15.6) 226(73.4)N+ (n = 107) 25(23.4) 17(15.9) 65(60.7)

Numbers in parenthesis are percentages.a The age of one patient was not recorded.

The clinical significance of the positive surgical margin in oral cancer 781

oral cancer including the T-stage of the tumor,6,8,15

N-stage,6,15 and histopathological features including tumorthickness4 and the pattern of invasion.6,11,13 These observa-tions indicate that the interactions between surgical marginstatus and treatment outcome in oral cancer are complex.This study was done to examine the independent effect ofthe status of the surgical margin on permanent pathologicsection on local recurrence and survival.

Methods

An historical cohort of 707 patients with squamous cell car-cinoma of the oral cavity from the cancer registry of theProvince of Manitoba, Canada from January 1975 to presentwas examined. After excluding patients with incomplete re-cords (n = 34), seen in consultation only or treated with pal-liative intent (n = 69), treated with radiotherapy as a singletreatment modality (n = 179) there were 425 cases of biopsyproven and previously untreated squamous cell carcinomaof the oral cavity managed with surgery with or withoutadjunctive radiotherapy. Pathology reports were reviewedfor the status of the resection margin. The presence of tu-mor at the inked resection margin was considered as posi-tive. Tumor identified within 2 mm of the inked resectionmargin was recorded as close. Tumors were staged usingthe TNM classification.16 Categorical data was evaluatedusing Pearson chi-square, with continuity correction whenappropriate, or Fisher’s exact test. A t-test or Mann–Whit-ney U-test was used for continuous variables. For survivalanalysis Kaplan–Meier curves17 and log-rank test for com-paring sub-groups was used. A multivariate Cox’s propor-tional hazard model18 was used to test the hypothesis ofincreased hazard or risk of death with close and involvedmargins after controlling for potentially confounding vari-ables, such as age, tumor stage, etc. In addition, this modelproduced the adjusted hazard ratios (HR)19 and the esti-mated adjusted probabilities of overall survival for eachpredictor in the model.

Results

The mean age of the cohort was 63.6 ± 12.6 years and58.89% were males. Tobacco or alcohol use or a combina-tion of both was documented in 80% of patients. One hun-dred and ninety-eight -second primary malignancies wereidentified in 145 patients, of which 61% were metachranous.The tongue was the primary site in 129 (30.4%) of patients,floor of mouth in 169 (39.8%), lower alveolus in 38 (8.9%),buccal mucosa in 33 (7.8%), upper alveolus in 23 (5.4%),and retromolar trigone in 33 (7.8%). One hundred andtwenty-four patients had Stage I lesions, while 118, 68,and 105 patients had Stage II, III, and IV disease, respec-tively. Ten patients could not be staged. One hundred andseven patients (26%) had clinically positive nodes at thetime of presentation.

Surgery was used as a single treatment modality in 298(70.1%) patients and 127(29.9%) received adjuvant radio-therapy. Per-oral excision of the primary tumor was per-formed in 243 patients and 182 were managed with acomposite resection. Mandible was resected in 175 patients,a marginal resection in 62 and segmental in 113. The clini-

cally negative neck was treated electively in 120 patients(39.0%). Adjunctive radiotherapy was delivered by conven-tional fractionation, to a median dose of 5000 CGy (mode5000 CGy). There were four deaths within 30 days of treat-ment completion. Minor morbidity was recorded in 23patients treated with surgery (7.7%) and 42 (33.1%) treatedwith adjunctive radiotherapy (P < 0.0001). Major morbiditywas observed in 8 (2.7%) of patients treated with surgeryand in 10 (7.9%) of patients receiving radiotherapy(P = 0.0313).

Sixty-two patients (14.6%) had microscopic tumor identi-fied at the inked resection margin, 68 (16.0%) patients hadclose (<2 mm), and 295 (69.4%) patients had clear (>2 mmmargins). The relationships between the status of the surgi-cal margins and clinical data is shown in Table 1. Involvedmargins were observed less frequently following excisionof tongue tumors, however the observed differences didnot achieve statistical significance (P = 0.1239). Involvedmargins were associated with both advanced T-stage(P = 0.0033) and the presence of clinically positive lymphnodes (P = 0.0034). The type of surgical resection did notinfluence the incidence of involved margins (Table 2). Theuse of intra-operative frozen section was not associatedwith either involved or clear margins. Adjunctive radiother-apy was used more frequently in patients with involved mar-gins (P = 0.0005).

Recurrent disease, observed most frequently at the pri-mary site and neck, was observed in 78 (43%) of patientstreated with surgery and 46 (36%) of patients treated withadjunctive radiotherapy (P = 0.24). The relationships be-tween margin status and the sites and incidence of recur-rent disease and overall survival is shown in Table 3.

Figure 1 Kaplan–Meier survival curves showing five-yearoverall survival for patients with involved, close, and clearmargins.

782 A. Binahmed et al.

Patients with involved and close margins had a similar inci-dence of treatment failure, which was significantly higherthan that observed in patients with clear margins. Failureat the primary site was more frequent in the presence ofboth involved and close margins. Margin status was not asso-ciated with failure in the neck or distant metastases. Onehundred and ninety-eight deaths were documented, 73attributed to oral cancer, 51 to second primaries, 44 fromother conditions, and could not be accurately determinedin 30. Sixty-two percent of the cohort survived five-years.Overall survival was significantly worse for patients with in-volved margins when compared to those with close or clearmargins (Fig. 1).

The Cox’s proportional hazard model is shown in Table 4.Age, T-stage, N-stage, and involved margins had a signifi-cant impact on five-year survival. After controlling for theconfounding variables the presence of tumor at the inkedresection margin increased the risk of death by 90% (HR1.9). The presence of a close margin did not have a signifi-cant impact on survival.

Table 2 Relationship of the status of the surgical margin totreatment parameters

Variable Involved(n = 62)

Close(n = 68)

Clear(n = 295)

ResectionPer-oral (n = 243) 31(12.8) 40(16.5) 172(70.7)Composite (n = 182) 31(17.1) 28(15.3) 123(67.6)

MandibulectomyMarginal (n = 62) 8(12.9) 14(22.6) 40(64.5)Segmental (n = 113) 23(20.3) 15(13.3) 75(66.4)None =(n = 250) 31(12.4) 39(15.6) 180(72)

Frozen sectionYes (n = 225) 36(16.0) 31(13.8) 158(70.2)No (n = 200) 26(13.0) 37(18.5) 137(68.5)

Adjunctive radiotherapySurgery (n = 298) 31(10.4) 47(15.8) 220(73.8)Surgery + radiotherapy(n = 127)

31(24.4) 21(16.5) 75(59.1)

Numbers in parenthesis are percentages.

Table 3 Relationship of the status of the surgical margin to theyear overall survival

Involved (n = 62) Close (n

Recurrence 25(40.3) 26(38.2)

Site of recurrencePrimary site 17(27.4) 16(23.5)Neck 14(22.6) 11(16.2)Distant Metastases 3(4.8) 5(7.4)

Overall survival at five-years 38.7% 58.3%

Numbers in parenthesis are percentages.

Table 4 Cox’s proportional hazard model predicting five-year overall survival

Variable Hazardratio

95% CI P-value

Included in the modelInvolved versus clear/

close margin1.9 (1.2, 2.9) 0.0026

Age P 65 versus Age < 65 1.5 (1.1, 2.2) 0.0017T-Status 1.4 (1.2, 1.6) 0.0002N-Status 1.3 (1.0, 1.6) 0.0465

Excluded from the modelClose versus clear margin 0.1027Gender 0.3212Treatment modality 0.6105Stage 0.8805Site 0.8379

Discussion

This historical cohort study of patients with squamous cellcarcinoma of the oral cavity is relatively large. It is compa-

incidence of recurrence, sites of initial recurrence, and five-

= 68) Clear (n = 295) All (n = 425) P-value

73(24.7) 124(29.2) P = 0.009

41(13.9) 74(17.4) P = 0.00537(12.5) 62(14.6) P = 0.11612(4.1) 20(4.7) P = 0.207

68.4% 61.9% P < 0.0001

The clinical significance of the positive surgical margin in oral cancer 783

rable in most respects, including the age and sex distribu-tion, site distribution, etiologic factors, and incidence ofsecond primaries, to most series of oral cancer.The inci-dence of involved margins (microscopic tumor at the inkedresection margin) observed in this population based histori-cal cohort was 14.6%. The reported incidence of involvedmargins following resection of head and neck cancers varieswidely from 1 to 22%.1–3,5–7,9,13,20 These studies with oneexception7 report single institutional experience and ofteninclude extra-oral sites.1,3,5,7 Margins are more often posi-tive in oral cancer than other sites in the head and neck.5,15

Tumor at the inked resection margin increased recur-rence rates in this study consistent with previous observa-tions.1,2,6,7,9 An independent and significant impact onoverall survival was demonstrated. Involved margins wereassociated with an increased tumor load, advanced T- andN-status, on presentation. The association of involved mar-gins with factors that can independently influence outcomein oral cancer such as stage of disease6,8,15 and adverse his-topathological features such as tumor thickness4 and unfa-vourable patterns of invasion6,11,13 make the use ofmultivariate models necessary to assess the complex inter-actions of margin status and survival. Sutton and co-work-ers,6 in a detailed analysis of the impact of margin statuson outcome, included clinical findings as well as histopa-thology data including nodal involvement and extracapsularspread, pattern of invasion, perineural invasion and vascularpermeation in their multivariate model. The relative risk ofdeath for involved margins was 11.61 (P = 0.0013) and closemargins 2.66 (P = 0.02). The only other factor with a signif-icant impact on survival was positive lymph nodes (HR 2.15,P = 0.063). Other studies have failed to show an indepen-dent effect of positive or inadequate margins on survivalwith multivariate analysis.5,13,14 Theses studies are notstrictly comparable to the above series as they included sig-nificant numbers of patients with extra-oral tumors5,14 pre-viously treated patients5 and a different definition for the‘‘positive margin’’.13,14

Close margins (<2 mm) had a similar impact on the inci-dence and pattern of local recurrence as involved margins.Survival was not adversely effected by the close margin.Many investigators consider both close, and tumor at theinked resection margins together in their analysis of the im-pact of margin status on outcome.8,11,21,22 When the impactof the close margin is reported separately the results sug-gest that the prognosis is better than observed in patientswith tumor at the resection margin.2,3,6,12 A major problemin assessing and comparing the results of studies evaluatingmargins is the fact that some series include patients withclose margins and others do not. This problem is furthercompounded by the lack of a standard definition for theclose margin. The most widely accepted definition of a closemargin is tumor within 5 mm of the inked resection mar-gin.1–3,6,8,23 This is an arbitrary designation.2,24 Others use<2 mm,13,25 <3 mm,12 within one high power field9 or a var-iable definition related to the pattern of invasion at the tu-mor host interface.10 In situ change and dysplasia have alsobeen included.1–3,6,8,21 This study did not address the im-pact of premalignant change at the margin. The balanceof opinion would favor not including dysplasia in the defini-tion.23 The results of this study suggest that the outcomewith tumor close to the resection margin should be reported

separately from that observed with tumor at the inkedresection margin as the prognosis differs. A standard defini-tion for a close margin is necessary and this should be deter-mined by a systematic evaluation of the effect of the widthof resection on outcome.

Considering the significant impact on outcome precisedefinitions for the inadequate surgical margin are importantclinically. Patients with assumed inadequate margins areconsidered for adjunctive radiotherapy. It is accepted thatadjunctive radiotherapy in the setting of an inadequatemay improve local-regional control26–28 and possibly impactsurvival.20,21 Adjunctive radiotherapy did not impact localand regional recurrence or survival in this study. It did how-ever significantly increase morbidity.

The surgical margin is a prognostic determinant that is,at least in part, under the control of the surgeon as he orshe determines the extent of resection in the individualpatient. The clinical margin necessary to achieve an ade-quate histological margin has to take into considerationthe fact that surgical specimens shrink, in the order of40–50%24 when fixed in formalin. Mucosal margins tendto receive the most attention, however most recurrencesinvolve the deep resection margin. The importance of athree-dimensional resection is emphasized in the work ofRavasz et al.8 Recurrence was not identified in patientswith positive mucosal margins. Positive deep and multiplepositive margins were associated with recurrence rates of38% and 70%, respectively. The positive association of mar-gin status with various tumor related factors observed inthis and other studies indicates that negative margins areharder to achieve with large tumor loads, proximity tomandible and a posterior location in the oral cavity. Thepositive association of margins with unfavorable histopa-thology such as tumor invading deeply as nests and cordsas opposed to a pushing front has prompted several inves-tigators6,14 to consider positive margins as an indicator ofmore aggressive disease. Patients with poor histopatholo-gical parameters with clear margins have a better outcomethan patients with bad histology and positive margins.8,22

The patient’s expectations for a functional outcome alsoinfluence the extent of resection. There is no single defini-tion for an adequate clinical resection margin. As definedby Yuen25 the optimal resection margin should not compro-mise local control from an inadequate resection or causeunnecessary functional morbidity from too much resection.The use of intra-operative frozen section did not influencemargin status in this study. Frozen section tends to empha-size mucosal and not the deep resection margin. In addi-tion the common practice of assessing the patientsresection margins,23 does not identify the patient with aclose margin.

In summary the status of the surgical resection is animportant predictor of outcome, both local recurrenceand overall survival in oral cancer. The presence of tumorat the resection margin, and close to the resection marginshould be considered separately with respect to prognosticsignificance. Intra-operative assessment of resection mar-gins needs to emphasize involvement and proximity of tu-mor to the deep resection margin. There is a need for astandard definition for the close margin in oral cancer basedon a systematic evaluation of the impact of the width of themargin on outcome.

784 A. Binahmed et al.

Conflict of interest statement

We the authors of this manuscript (A. Binahmed, R. Nason,A. Abdoh), have no financial or personal relationships withother people or organizations that could inappropriatelyinfluence our work.

References

1. Looser KG, Shah JP, Strong EW. The significance of ‘‘positive’’margins in surgically resected epidermoid carcinomas. HeadNeck 1978;1:107–11.

2. Loree TR, Strong EW. Significance of positive margins in oralcavity squamous carcinoma. Am J Surg 1990;160:410–4.

3. Chen TY, Emrich LJ, Driscoll DL. The clinical significance ofpathological findings in surgically resected margins of theprimary tumor in head and neck carcinoma. Int J Radiat OncolBiol Phys 1987;3(6):833–7.

4. Yuen APW, Lam KY, Wei WI, et al. A comparison of theprognostic significance of tumor diameter, length, width,thickness, area, volume, and clinicopathological features oforal tongue carcinoma. Am J Surg 2000;180:139–43.

5. Jones AS. Prognosis in mouth cancer: tumour factors. OralOncol Eur J Cancer 1994;30B(1):8115.

6. Sutton DN, Brown JS, Rogers SN, Vaughan ED, Woolgar JA. Theprognostic implications of the surgical margin in oral squamouscell carcinoma. Int J Oral Maxillofac Surg 2003;32:30–4.

7. Jacobs JR, Ahmad K, Casiano R, et al. Implications of positivesurgical margins. Laryngoscope 1993;103:64–8.

8. Ravasz LA, Slootweg PJ, Hordijk GJ, Smit F, van der Tweel I.The status of the resection margin as a prognostic factor in thetreatment of head and neck carcinoma. J Cranio Max Fac Surg1991;19:314–8.

9. Spiro RH, Guillamondegui Jr O, Paulino AF, Huvos AG. Patternof invasion and margin assessment in patients with oral tonguecancer. Head Neck 1999;21:408–13.

10. Slootweg PJ, Hordijk GJ, Schade, van Es RJJ, Koole R.Treatment failure and margin status in head and neck cancer.A critical view on the potential value of molecular pathology.Oral Oncology 2002;38:500–3.

11. Woolgar JA, Brown JS, Scott J, West CR, Vaughan ED,Rogers S. Survival, metastasis and recurrence of oral cancerin relation to pathological features. Ann R Coll Surg Engl1995;77:325–31.

12. Chandu A, Adams G, Smith ACH. Factors affecting survival inpatients with oral cancer: an Australian perspective. Int J OralMaxillofac Surg 2005;34:514–20.

13. Kademani D, Bell RB, Bagberi S, Homgren E, Dierks E, Potter B,et al. Prognostic factors in intraoral squamous cell carcinoma:

The influence of histologic grade. J Oral Maxiollofac Surg2005;63:1599–605.

14. McMahon J, O’Brien CJ, Pathak I, et al. Influence of conditionof surgical margins on local recurrence and disease-specificsurvival in oral and oropharyngeal cancer. Br J Oral MaxiollofacSurg 2003;41:224–31.

15. Woolgar JA, Triantafyllou A. A histopathological appraisal ofsurgical margins in oral and oropharyngeal cancer resectionspecimens. Oral Oncology 2005;41:1034–43.

16. International Union Against Cancer. TNM Classification ofMalignant Tumors. sixth ed. New York: Wiley-Liss; 2002., pp.22–6.

17. Kaplan ER, Meier P. Non-parametric estimation from incom-plete observations. J Am Stat Assoc 1958;53:457–81.

18. Cox DR. Regression models and life tables. J R Stat Soc1972;34:187–9.

19. Spruance SL, Reid JE, Grace M, Samore M. Hazard ratio inclinical trials. Antimicrob Agents Chemother 2004;48(8):2787–92.

20. Sessions DG, Spector GJ, Lenox J, Haughey B, Chao C, Marks J.Analysis of treatment results for oral tongue cancer. Laryngo-scope 2002;112:616–25.

21. Sadeghi A, Kuisk H, Tran LM, Mackintosh R, Mclaren JR, ParkerRG. The role of radiation therapy in squamous cell carcinoma ofthe upper aerodigestive tract with positive surgical margins.Am J Clin Oncol (CCT) 1986;9(6):500–3.

22. van Es RJJ, van Nieuw Amerongen N, Slootweg PJ, Egyedi P.Resection margin as a predictor of recurrence at the primarysite for T1 and T2 oral cancers. Arch Otolaryngol Head NeckSurg 1996;122:521–5.

23. Meier JD, Oliver DA, Varvares MA. Surgical margin determina-tion in head and neck oncology: Current clinical practice. Theresults of an international American head and neck societymember survey. Head Neck 2005;27:952–8.

24. Batsakis JG. Surgical excision margins: A pathologist’s perspec-tive. Adv Anatomic Pathol 1999;6(3):140–8.

25. Yuen PW, Lam KY, Chan ACL, Wei WI, Lam LK. Clinicopatho-logical analysis of local spread of carcinoma of the tongue. AmJ Surg 1998;175:242–4.

26. Vikram B, Strong EW, Shah JP, Spiro R. Failure at the primarysite following multimodality treatment in advanced head andneck cancer. Head Neck Surg 1984;6(3):720–3.

27. Zelefsky MJ, Harrison LB, Fass DE, Armstrong JG, Shah JP,Strong EW. Postoperative radiation therapy for squamous cellcarcinomas of the oral cavity and oropharynx: Impact oftherapy on patients with positive surgical margins. Int JRadiation Oncology Biol Phys 1992;25:17–21.

28. Hinerman RW, Mendenhall WM, Morris CG, Amdur RJ, WerningDB, Villaret DB. Postoperative irradiation for squamous cellcarcinoma of the oral cavity: 35-year experience. Head Neck2004;26:984–94.